Hospital services utilisation and cost before and after COVID-19 hospital treatment: Evidence from Indonesia

Muhammad Fikru Rizal; Firdaus Hafidz; Gilbert Renardi Kusila; Wan Aisyiah; Dedy Revelino; Erzan Dhanalvin; Ayunda Oktavia; Ilyasa; Citra Jaya; Benjamin Saut; Mahlil Ruby

doi:10.1371/journal.pone.0305835

. 2024 Jul 5;19(7):e0305835. doi: 10.1371/journal.pone.0305835

Hospital services utilisation and cost before and after COVID-19 hospital treatment: Evidence from Indonesia

Muhammad Fikru Rizal ^1,², Firdaus Hafidz ^1,^*, Gilbert Renardi Kusila ¹, Wan Aisyiah ³, Dedy Revelino ³, Erzan Dhanalvin ³, Ayunda Oktavia ³, Ilyasa ³, Citra Jaya ³, Benjamin Saut ³, Mahlil Ruby ³

Editor: Nemer Badwan⁴

PMCID: PMC11226039 PMID: 38968247

Abstract

Objective

To estimate hospital services utilisation and cost among the Indonesian population enrolled in the National Health Insurance (NHI) program before and after COVID-19 hospital treatment.

Methods

28,159 Indonesian NHI enrolees treated with laboratory-confirmed COVID-19 in hospitals between May and August 2020 were compared to 8,995 individuals never diagnosed with COVID-19 in 2020. A difference-in-difference approach is used to contrast the monthly all-cause utilisation rate and total claims of hospital services between these two groups. A period of nine months before and three to six months after hospital treatment were included in the analysis.

Results

A substantial short-term increase in hospital services utilisation and cost before and after COVID-19 treatment was observed. Using the fifth month before treatment as the reference period, we observed an increased outpatient visits rate in 1–3 calendar months before and up to 2–4 months after treatment (p<0.001) among the COVID-19 group compared to the comparison group. We also found a higher admissions rate in 1–2 months before and one month after treatment (p<0.001). Consequently, increased hospital costs were observed in 1–3 calendar months before and 1–4 calendar months after the treatment (p<0.001). The elevated hospital resource utilisation was more prominent among individuals older than 40. Overall, no substantial increase in hospital outpatient visits, admissions, and costs beyond four months after and five months before COVID-19 treatment.

Conclusion

Individuals with COVID-19 who required hospital treatment had considerably higher healthcare resource utilisation in the short-term, before and after the treatment. These findings indicated that the total cost of treating COVID-19 patients might include the pre- and post-acute period.

Introduction

As the COVID-19 pandemic continues to become a global health issue, the understanding of its consequences beyond the respiratory system and the acute treatment for the disease itself has become a major concern. Several studies found that COVID-19 infection does not only affect the respiratory system but also haematological, cardiovascular, neurological, and gastrointestinal organs [1]. It has also been found that older people and those with chronic illnesses are more vulnerable to developing severe and critical COVID-19 infections. Moreover, these groups also have a higher mortality rate [2]. A study conducted in the United States found that COVID-19 hospitalisation was six times higher in patients with pre-existing comorbidities than in those without [3]. In Indonesia, about 12% of the hospitalised patients died in the first three months of the pandemic, with higher rate among older people, male, and those with multiple comorbidities [4]. A more recent study indicated that districts with higher proportion of elderly and lower healthcare capacity had higher rate of COVID-19 mortality [5].

Apart from our increasing knowledge related to the risk factors for COVID-19 and its mortality, our understanding of the clinical sequelae of COVID-19 infections has grown over time. Due to its multi-organ involvement, some COVID-19 symptoms can persist for up to six months after infection remission or more, which is defined as “long COVID” or “post-COVID” syndrome [6]. A meta-analysis conducted by Chen et al. found that fatigue, memory problem, and respiratory issue are the three most prevalent symptoms [7]. Current evidence indicates that older patients, those with more than five symptoms and hospitalised, and the Asian population are also identified to be more likely to experience long covid [6–8].

The daily activities and quality of life deteriorate among the people who suffer from long covid, with physical limitations and emotional stress included [9, 10]. The post-covid examination and therapies are then needed in a large proportion of recovered patients. Furthermore, routine pulmonary rehabilitation and some medication consumptions were marked effective in decreasing dyspnoea among long covid patients [11]. From a policymaker’s perspective, the importance of long-covid cannot be understated. In the short run, the health system in each country needs to anticipate the potential change in healthcare needs through better allocation of resources and investment [12]. However, evidence on the potential impact of COVID-19 infections on subsequent healthcare use and costs is scarce, especially in low-resource settings.

Findings from National Health Services (NHS) in England indicate that COVID-19 hospitalisation increases the readmission rate and subsequent diagnosis of respiratory diseases, diabetes, and cardiovascular diseases [13]. For mild infections, a study in Norway found a short-term elevation of primary care use, mostly for a respiratory and general or unspecified condition, but no effect on specialist care. This change, however, mostly reverted after three months [14, 15]. While these studies used a large administrative database, hence minimising reporting bias, none of them reported the costs associated with pre- or post-COVID changes in healthcare use.

This paper aimed to fill the literature gap by estimating the change in utilisation and cost of hospital services before and after individuals were treated in hospitals with laboratory-confirmed COVID-19. We leveraged administrative National Health Insurance (NHI) data from Indonesia, a lower-middle-income country (LMIC) with about 260 million population. The Indonesian NHI is the biggest single-payer social health insurance program in the world, covering more than 90% of the Indonesian population [16]. Our study aims to examine the impact of COVID-19 by analysing both the direct effects on patients treated for the disease and the associated changes in hospital service utilization and costs. This dual perspective provides insights into the short- to medium-term adjustments within hospital services pre- and post-acute COVID-19 treatment periods, especially critical for understanding the pandemic’s ramifications in resource-constrained settings.

Methods

Data source

We utilised all COVID-19-related hospital claims which were recorded in the deidentified National Health Insurance (NHI) database from January 2020 to February 2021. Patients who were in this COVID-19 claims data were linked with their regular hospital claims records dated back to December 2016. The merged dataset contains information related to patients’ basic demographic characteristics and healthcare utilisation records. The basic demographic characteristics include the year of birth, sex, province where individuals were treated with COVID-19 and type of NHI membership (subsidised or non-subsidised scheme).

The utilisation records contain the month of encounter, type of services (inpatient or outpatient), primary and secondary diagnoses as comorbidity, diagnoses-related group codes (used as the basis for hospital payment), severity level, levels of hospitals (a basic type D hospital to an advanced type A hospital), amount of claims, and the discharge status (in-hospital death, routine, or against medical advice), and COVID-19 diagnosis status (suspect, probable, or laboratory-confirmed).

Ethical clearance

Our study did not obtain the written consent from the participants, as it used an NHI database as it source. All data were fully anonymized before being accessed. Ethical clearance was obtained from the Ethical Committee at Faculty of Medicine, Public Health, and Nursing, Universitas Gadjah Mada (No KE/FK/0945/EC/2021).

Study population

Individuals who were diagnosed based on positive RT-PCR results and treated, either in outpatient care or admitted to inpatient care, from May to August 2020 were included in the COVID-19 group. Individuals who were under monitoring and recorded as suspect or probable cases of COVID-19 in the previous months before laboratory confirmation were excluded. Similarly, those who died during COVID-19 hospitalisation or in subsequent months were also excluded from the analysis. This decision was made to focus on COVID-19 survivors, hence avoiding potential issue of reduced healthcare use of the COVID-19 group due to their higher post-hospital treatment mortality. For context, the recorded 6-month mortality rate of those treated with COVID-19 between May and August 2020 was 12.1%. The comparison group comprised individuals treated for COVID-19 in hospitals in February 2021 (not hospitalised in 2020), under the assumption they had not contracted the virus in 2020. We argue this is a suitable comparator for our analysis as these groups are similar (i.e. attitude towards risk, preference, etc.). As we observed their health care utilisation in the entirety of our estimation sample from August 2019 to November 2020, the event study approach allows us to compare the outcomes between treatment and comparison groups over time. This assumption is plausible given reinfection rate within one year follow up was only about 5% [17]. Note that in the NHI database, only results from RT-PCR (Reverse Transcriptase Polymerase Chain Reaction) done by Ministry of Health-accredited laboratories are used [18].

Study period

We considered a period of August 2019 to November 2020 in our analysis. During this period, the comparison group, those who were treated with COVID-19 in February 2021, was assumed to not yet contracted the disease. December 2020 and January 2021 were excluded due to the possibility of elevated healthcare utilisation two months before individuals in the comparison group were hospitalised with COVID-19 in February 2021. Hence, we had a maximum of six months (June 2020 to November 2020) and a minimum of three months (September to November 2020) post-COVID-19 follow-up. For pre-COVID period, we included nine months before COVID-19 hospital treatment. The grouping strategy and the period of observation used in our analyses is shown in Fig 1.

Outcome variables

We measured the total number of hospital service utilisation, both inpatient and outpatient, and the cost of direct healthcare-related services as total reimbursement from national health insurance at the hospital (in Indonesian Rupiahs) separately for COVID-19 diagnosis and other diagnoses. Our estimates were presented in thousand Indonesian Rupiah (US$1 equals to Rp14,500). All these measures were recorded for each individual on a monthly basis. During the month of COVID-19 diagnosis and treatment, all the patients would have at least one hospital encounter.

Statistical analysis

Baseline characteristics of the COVID-19 group (overall and separately by the month of treatment) and the comparison group were described as percentages (for categorical variables) or mean and standard deviation (for continuous variable).

Next, we employed a difference-in-difference (DiD) event-study approach as implemented by Skyrud et al. and Magnusson et al. to estimate the association between COVID-19 hospital treatment and the utilisation and costs of hospital services, before and after the treatment [14, 15]. This technique estimated the association between an exposure and an outcome by contrasting the change in outcome among the exposed group from a given time before exposure (the reference period) to another time point with the change in outcomes among the comparison group within the same time frame. To identify causal effect of an exposure, it is assumed that in the absence of exposure, there should be no difference in outcomes trends over time between the exposed (COVID-19) group and the unexposed (comparison) group [19, 20].

Empirically, we estimated the following regression model (Eq 1).

Y_{i t} = Σ_{t = M i n}^{t = M a x} {β_{1}}_{t} M o n t h_{t} \times C o v i d_{i} + β_{2} C o v i d_{i} + β_{3} M o n t h_{t} + β_{4} X_{i} M o n t h_{t} + β_{5} X_{i} + ε_{i t}

(1)

The Y_it variable represents the outcomes of individual i in month t. Month_t is a categorical variable indicating the month or period of observation, with month zero indicating the month where individual was treated for laboratory-confirmed COVID-19 infection. Min and Max indicate how many months (or groups of months) before and after the month of hospitalisation, respectively (see Fig 1 for how we constructed these observations period). For hospital service utilisation regressions, we used the fifth month leading to the hospitalisation as the reference period. For example, May 2020 group, this reference period was December 2019 and labelled as t-5. The t0 was May 2020, the t+1 to t+6 was June to November 2020, the t-4 to t-1 was January 2020 to April 2020. For each subsequent month, the reference and comparative periods shifted accordingly”.

As for cost regression, we used the period of 6 to 4 months before hospitalisation as the reference period and lumped together 9 to 7 months before the hospitalisation as pre-reference period. Covid_i is a binary variable indicating whether individuals are in the COVID-19 or in the comparison group. As previously discussed, nine months period before exposure and three to six months period after exposure were used in the main analyses. Coefficient ${β_{1}}_{t}$ capture the month specific DiD estimates measuring the association between COVID-19 hospital treatment and each of the outcomes.

Although some of our outcomes were count variables, linear modelling can still be an appropriate choice [21], hence we estimate Eq 1 using Ordinary Least Square (OLS). Next, to account for differences in characteristics between COVID-19 and comparison group, we also include set of control variables (X_i) which include age, sex, severity, comorbidities diagnosed before January 2020, NHI membership segment, and province. These variables are interacted with month indicator to allow for their changing influence on outcomes over time. Severity levels in our study are defined based on the Indonesian Case-Based Groups (INA-CBGs) system, a case-mix payment system that utilizes a software grouper application. The severity level is influenced by complications and comorbidities, which are indicative of the resource intensity level required for treatment during the first treatment episode where patients were laboratory-confirmed. Heterogeneity analysis was conducted by splitting the sample into younger (<40 years old) and older (≥40) population, as well as by gender. The age cut-off was considered given the risk of major comorbidities such as cardiovascular diseases and cancer known to increase exponentially after the age of 40 [22].

Results

Study participants

Of 247,650 NHI members with COVID-19-related hospital encounters between January 2020 and February 2021, 146,240 were confirmed to be infected by COVID-19 based on their RT-PCR results, while the others were categorised as suspect or probable cases only. We excluded those who died during or after COVID-19 treatment and those who were hospitalised as suspect or probable COVID-19 cases in the previous month before laboratory confirmation. Fig 2 depicts the flow of study participants and their classification into COVID-19 and comparison group. Our final sample consisted of 28,159 individuals in the COVID-19 group (those in hospitals with COVID-19 between May and August 2020) and 8,995 individuals in the comparison group (those in hospitals with COVID-19 in February 2021).

The average age of our sample (per January 2020) ranged from 44.2 to 46.5 years old, with some variations between groups. There were slightly more females (51.7% to 54.2%) and substantially higher individuals who were non-subsidised members of the NHI (81% to 85.9%). The subsidised scheme of the NHI, either provided by the central or local government, was intended to target the poorer population. Based on the distribution of severity level, which reflects the intensity of hospital resource utilization, most of the patients were hospitalized when they were confirmed with COVID-19 infection, although with decreasing trend (96% in May 2020 and 89% in the comparison group). Most of the study population were from Java and Bali provinces (58.2% for May 2020 group to 80.7% in the comparison group). Lastly, hypertension (11.1% to 14.6%) and diabetes (10.9% to 13%) were the two most common pre-COVID-19 comorbid–defined as diagnoses received by individuals before January 2020 (see Table 1).

Table 1. Baseline characteristics of COVID-19 cases treated in hospitals in Indonesia from May to August 2020 compared with the comparison group.

	COVID-19 group–Month in hospital with COVID-19					Comparison group
	May 20	Jun. 20	Jul. 20	Aug. 20	All	Comparison group
Age, mean (SD)	44.6 (15.8)	44.5 (15.8)	44.2 (15.6)	45.4 (15.6)	44.7 (15.6)	46.5 (16.2)
Age group, %
<20	4.3	4.1	4.2	3.8	4.0	4.3
20–39	34.8	34.6	34.7	32.4	33.9	29.9
40–59	42.3	42.4	44.5	45.1	44.1	42.9
≥40	18.6	18.8	16.6	18.8	18.1	22.9
Sex, %
Male	46.9	46.4	48.3	48.3	47.8	45.8
Female	53.1	53.6	51.7	51.7	52.2	54.2
Membership scheme, %
Subsidised	19.0	18.3	16.2	17.0	17.2	14.1
Non-subsidized	81.0	81.7	83.8	83.0	82.8	85.9
Severity, %
None (outpatient)	4.4	7.0	7.4	9.2	7.7	10.7
Level 1	62.8	60.8	64.2	61.3	62.3	61.9
Level 2	11.4	13.1	13.3	14.4	13.5	12.3
Level 3	21.4	19.1	15.0	15.1	16.5	15.1
Comorbidity, %
Hypertension	11.1	12.7	11.6	12.7	12.2	14.6
Diabetes	10.9	11.8	11.2	12.3	11.7	13.0
Heart disease	6.2	7.2	6.4	7.4	6.9	8.2
Tuberculosis	2.3	2.1	1.8	1.8	1.9	2.3
Asthma	2.4	2.0	2.4	2.3	2.3	2.4
COPD	2.0	1.6	1.3	1.5	1.5	1.8
Cancer	4.1	4.4	3.9	4.4	4.2	4.6
Liver disease	1.8	1.4	1.3	1.3	1.4	1.6
Kidney failure	2.7	2.7	2.6	3.2	2.8	2.9
Provinces by island, %
Sumatera	13.8	9.2	9.6	12.6	11.0	8.7
Java-Bali	58.3	63.4	68.9	73.0	68.2	80.7
Kalimantan	7.3	7.7	7.5	7.4	7.5	4.6
Sulawesi	13.5	12.7	7.1	3.2	7.5	1.0
Nusa Tenggara, Maluku, and Papua	6.4	6.4	5.4	2.7	4.7	4.7
Observations	2,796	5,777	9,237	10,349	28,159	8,995

Open in a new tab

Note: COVID-19 group were divided into month based on their initial laboratory-confirmed diagnoses that were recorded in the COVID-19 hospital claims dataset. Individuals who died of any causes during or after COVID-19 treatment were excluded. Comparison group consisted of individuals who were diagnosed and treated in hospital with COVID-19 in February 2021. Individuals’ age was in January 2020. Severity level is based on hospital resource utilisation during the first treatment episode where patients were laboratory-confirmed.

Changes in utilisation of hospital services

Fig 3 presented the difference-in-difference event-study estimates in outpatient visits rate of the COVID-19 group compared with the comparison group separately by treatment month. We presented the results for all ages, by age (<40 and ≥40 years old), and by gender. Our results indicated that up to the reference month, the differences in outcome between the COVID-19 and comparison group were close to zero and all statistically insignificant. This finding supported our assumption that the trend in outcome between the two groups were similar in the absence of COVID-19.

Starting from three months before treatment month, we observed some statistically significant increases in outpatient utilisation rate that peaked at the month of treatment with about 20 more visits per 100 individuals per month (p-value <0.001). In the subsequent months, those in hospitals with COVID-19 from June to August were still using more outpatient visits in about two to three months. A similar pattern was not observed for those who got COVID-19 in May. Further analysis indicated that this post-COVID trend was driven mainly by the older population. There were no substantial differences between males and female prior to month of treatment. However, the post-treatment trend indicated that males used more outpatient services after treated with COVID-19 (Fig 3).

As for inpatient utilisation, overall, we only observed a significant increase of about 21 to 33 additional admissions per 1,000 individuals (p < 0.001) at one month before and after COVID-19. There is no apparent gap between younger and older individuals or males and females with regards to inpatient utilisation as shown by the overlapping confidence interval in Fig 4.

Fig 4 — Note: The outcome variable is the number of inpatient stays per 100 individuals per month. The horizontal axis represented months relative to COVID-19 hospital treatment. Lines represented the estimated difference between the COVID-19 and comparison groups, controlling for demographic characteristics. Shaded areas represented the 95% Confidence Intervals. Estimates for the month during laboratory-confirmed COVID-19 in hospital treatment were omitted because almost all individuals in the COVID-19 group had at least one hospital encounter.

Changes in total costs of hospital services

Tables 2–4 presented the estimated difference-in-difference event-study in total costs of hospital services of the COVID-19 group compared with the comparison group, separately for each month when individuals were in hospitals with COVID-19. Results for all ages were shown in Table 2, while heterogeneity analyses by age and gender were provided in Tables 3 and 4, respectively.

Table 2. Difference-in-difference estimates of total costs of hospital services of COVID-19 group compared to comparison group.

	Month in hospital with COVID-19
	May 20		Jun. 20		Jul. 20		Aug. 20
-7 to -9 month	-30.2	(31.6)	-21.8	(22.9)	-3.36	(21.1)	-24.7	(22.1)
-4 to -6 month	Ref.
-3 month	27.7	(42.7)	39.6	(32)	48.5*	(22.8)	3.76	(21.6)
-2 month	32.2	(44)	76.2**	(27.5)	22.8	(20.2)	-4.99	(21.8)
-1 month	129**	(45.3)	107**	(35.1)	126***	(35.6)	56.1*	(27.6)
0 (COVID-19)	168,649***	(2,259)	136,843***	(1,285)	109,998***	(860)	99,732***	(734)
+1 month	2,453***	(425)	2,377***	(313)	1,507***	(186)	1,071***	(116)
+2 month	611***	(178)	286***	(67.1)	296***	(65.1)	310***	(63.5)
+3 month	277*	(139)	357**	(111)	133**	(47.4)	102*	(41.8)
+4 month	457*	(206)	104*	(51.9)	165**	(50.1)
+5 month	84.1	(79.7)	224**	(73.6)
+6 month	76.4	(75.9)
Observations	188,640		221,565		255,206		251,459
Individuals	11,790		14,771		18,229		19,343

Open in a new tab

Note: Coefficients represented the estimated differences in total costs of hospital services (in thousand Rupiahs) for both outpatient and inpatient and for all causes or diagnoses. Regressions control for individuals’ year of birth, gender, COVID-19 severity, NHI membership segment, comorbidities prior to 2020 interacted with month indicator, as well as month and province fixed effects. Standard errors clustered at individual level in parentheses * p < 0.05, ** p < 0.01, *** p < 0.001.

Table 4. Difference-in-difference estimates of total costs of hospital services of COVID-19 group compared to comparison group by gender.

	Month in hospital with COVID-19
	May 20		Jun. 20		Jul. 20		Aug. 20
Panel A. Male
-7 to -9 month	-58.6	(56.3)	-34.9	(39.6)	1.82	(35.2)	-51.8	(38.3)
-4 to -6 month	Ref.
-3 month	11.2	(71)	24.4	(47.9)	52.3	(37.1)	45.9	(39.1)
-2 month	71.8	(81.1)	58.7	(41.3)	51.9	(32)	-29	(34.3)
-1 month	132	(84.5)	119	(60.9)	175**	(63.9)	19.9	(45.1)
0 (COVID-19)	173,020***	(3,373)	144,019***	(1,956)	116,517***	(1,277)	105,726***	(1,105)
+1 month	2,159***	(513)	2,294***	(444)	1,279***	(188)	1,186***	(184)
+2 month	718**	(251)	333**	(112)	333***	(100)	357***	(105)
+3 month	111	(104)	327*	(134)	92.1	(65.9)	165*	(73)
+4 month	256	(158)	3.1	(61.7)	217**	(76.8)
+5 month	112	(141)	305*	(136)
+6 month	168	(137)
Observations	86,832		101,985		120,064		118,534
Individuals	5,427		6,799		8,576		9,118
Panel B. Female
-7 to -9 month	-10.1	(34.2)	-10.1	(25.5)	-5.03	(24.7)	-.427	(23.9)
-4 to -6 month	Ref.
-3 month	38.5	(50.3)	52.8	(43)	48.1	(27.4)	-32.7	(21.6)
-2 month	-3.14	(44.9)	89*	(36.2)	-1.86	(25.7)	16	(27.1)
-1 month	129**	(43.7)	97.3*	(39.8)	78.2*	(34.2)	87.1*	(33.9)
0 (COVID-19)	164,843***	(3,050)	130,447***	(1,690)	104,157***	(1,160)	94,272***	(971)
+1 month	2,694***	(661)	2,423***	(429)	1,723***	(314)	962***	(145)
+2 month	522*	(258)	241**	(78.8)	270**	(85.5)	262***	(73.6)
+3 month	407	(241)	386*	(170)	173*	(68.1)	47.7	(45.4)
+4 month	627	(354)	189*	(80.6)	126	(66.9)
+5 month	61.5	(88.8)	162*	(72.3)
+6 month	-26.2	(67.8)
Observations	101,808		119,580		135,142		132,925
Individuals	6,363		7,972		9,653		10,225

Open in a new tab

Table 3. Difference-in-difference estimates of total costs of hospital services of COVID-19 group compared to comparison group by age.

	Month in hospital with COVID-19
	May 20		Jun. 20		Jul. 20		Aug. 20
Panel A. Age <40
-7 to -9 month	-19.7	(33.5)	32.7	(24.2)	39.2	(22.1)	-22	(21.1)
-4 to -6 month	Ref.
-3 month	-58.2	(46.7)	50.1	(29.6)	33	(30.7)	-2.23	(29.6)
-2 month	42.9	(48.8)	5.18	(34.3)	-3.75	(26.8)	11.9	(26.4)
-1 month	109	(79.8)	21.9	(27.6)	99.1	(54.5)	23.2	(32.4)
0 (COVID-19)	136,762***	(3,435)	107,241***	(1,792)	86,150***	(1,161)	76,387***	(984)
+1 month	1,197**	(369)	2,012***	(404)	1,207***	(207)	780***	(136)
+2 month	200	(166)	274*	(107)	289***	(83.9)	237**	(77.4)
+3 month	128	(115)	331*	(151)	167*	(78.8)	79.6	(61.1)
+4 month	309	(237)	57.4	(62.2)	132	(69.6)
+5 month	84.3	(153)	380*	(172)
+6 month	-51.2	(78.6)
Observations	66,736		79,710		93,338		88,582
Individuals	4,171		5,314		6,667		6,814
Panel B. Age ≥ 40
-7 to -9 month	-44.2	(48)	-53.2	(34)	-26.9	(30.9)	-22.3	(32.2)
-4 to -6 month	Ref.
-3 month	71.1	(62.9)	35.2	(47.5)	57.9	(31.5)	7.04	(29.2)
-2 month	21.7	(65.1)	117**	(38.8)	35	(28)	-15.4	(30.1)
-1 month	140*	(55.5)	159**	(53)	138**	(45.8)	71.5	(38.7)
0 (COVID-19)	186,400***	(2,910)	154,274***	(1,714)	123,675***	(1,153)	112,049***	(973)
+1 month	3,176***	(634)	2,599***	(440)	1,677***	(268)	1,231***	(163)
+2 month	861**	(267)	276***	(82.1)	291**	(89.1)	352***	(88.6)
+3 month	356	(207)	375*	(153)	105	(58.5)	116*	(55)
+4 month	546	(297)	138	(74.6)	175**	(67.1)
+5 month	89.3	(91.9)	136*	(58.3)
+6 month	155	(115)
Observations	121,904		141,855		161,868		162,877
Individuals	7,619		9,457		11,562		12,529

Open in a new tab

In Table 2, we found a statistically significant increase of about 107 to 129 thousand Rupiahs in monthly hospital costs one month before the month of COVID-19 treatment (p<0.01), except for those who were treated in August 2020 with only 56 thousand Rupiahs increase (p<0.05). In addition, those who were treated in June and July seemed to have higher monthly hospital costs starting from three months (for those treated in July) or two months (for those treated in June). Compared to the difference at the reference time (month -4 to -6), we found no statistically significant gap in inpatient use in the previous months (month -7 to -9) between COVID-19 and the comparison group. This finding indicated that the two groups were relatively similar beyond six months before COVID-19 treatment.

During the month of treatment, the NHI expensed about 100 to 169 million Rupiahs more to those with COVID-19 than the comparison group (p<0.001). The figure was higher for individuals aged 40 years or older, with about 112 to 186 million Rupiahs increase in expenditure (p<0.001). Similarly, the cost was also higher among males with COVID-19 with average cost of 106 to 173 million rupiahs more compared to the comparison group. Afterwards, we still observed greater monthly hospital claims up to five months, with a decreased amount over time. The differences ranged between 1 to 2.5 million rupiahs in the first month (p<0.001) to about 84 to 224 thousand rupiahs in the fifth month after COVID-19 treatment (although it is only statistically significant with p<0.01 for those who were treated in June 2020) (see Table 2).

Lastly, our analysis by age group revealed that the higher post-Covid hospital claims were concentrated in the older group (see Table 3). No substantial differences were found between males and females in post-Covid hospital claims (see Table 4).

In addition to the event-study estimates presented above, for transparency, we also reported the raw mean of monthly outpatient utilisation, inpatient utilisation, and hospital costs among COVID-19 groups and the comparison group (for all individuals) in Tables B1-B3 in S2 Table respectively.

Discussion

Main findings

Our results indicated some elevated uses and costs of hospital services months before and after the treatment month. Specifically, we found an increase in outpatient utilisation that happened three months before and up to three months after the treatment and an uptick in admissions rate one month before and after the treatment. These observed post-COVID trends were consistent with a previous study in England that found a higher readmission rate among discharged COVID-19 patients compared to comparison group [13]. Other studies reported that 20% of COVID-19 patients might require rehospitalisation, especially in older individuals and those with comorbidities [23–25]. This finding was consistent with our results where higher post-COVID utilisation was observed mainly in the older patients. In contrast, two studies in Norway and Denmark that only included mild cases did not find a short-term increase in specialist care [14, 15, 26].

Although multiple studies have reported that the presence of sequelae after SARS-CoV-2 infection might explain the elevated healthcare use in the post-treatment period [13, 27], other factors might still exist. The example includes the exacerbation of prior comorbidities or newly diagnosed conditions caused by extensive examinations during patients’ acute COVID-19 treatment episodes. For the latter, our findings of a short-term increase in utilisation indicated that, if this situation happened, the conditions were unlikely to be chronic.

Another important result from this study was the increased hospital utilisation before the treatment month. One possibility is that people with recent medical conditions were more likely to have more severe symptoms, hence increasing the risk of requiring hospital treatment [28]. The second one is the limited testing capacity and clear clinical guide in the early phase of the pandemic in Indonesia. This situation could lead individuals to visit healthcare facilities multiple times before being treated as confirmed COVID-19 cases [29–31]. In addition, the increased use of healthcare and costs from five months before treatment could suggest that patients with pre-existing conditions were more likely to experience severe COVID-19 outcomes, necessitating more extensive healthcare utilization [32].

This study also found a substantial variation with respect to the estimated changes in hospital costs across individuals treated in different calendar months. For example, during the month of treatment, patients treated in May 2020, on average, spent 169 million Rupiahs more than the comparison group. This figure decreased consistently with an estimated difference of about 138 million, 108 million, and 98 million Rupiahs among those treated in June, July, and August, respectively. Broadly, COVID-related claims in Indonesia during this period were based on a fixed per day payment scheme with values depending on the type of services (i.e. requiring ICU with or without a ventilator, negative pressure isolation room, or regular isolation room) and patient’s comorbidities. In addition, the government also reimbursed hospital expenditure on personal protective equipment (PPE) [33, 34]. Therefore, any gaps in costs between individuals treated in different calendar months might reflect changes in hospital resource usage. One plausible reason for this pattern is the rapid development of clinical guidelines for COVID-19 management [35].

Topics for future study

Given our results, a more detailed analysis of the type of diagnosis and procedure, including the use of primary care, is needed to fully understand the specific cause and type of increased healthcare utilisation in Indonesia. This type of question requires linked hospital and primary care utilisation, similar to what have been done in high-income countries [13, 14, 26, 27, 36]. The expanded database could also be utilised to examine the impact of less severe case of COVID-19 on healthcare utilisation.

In addition, since our study periods are mostly before the mass COVID-19 vaccination [37], a study using more recent data and a vaccination record is also needed. The observed patterns might change since the vaccination program is known to affect the course of COVID-19 disease. Lastly, due to the rapidly changing variants with different disease characteristics, data from different periods will be highly valuable in exploring the heterogeneity of its impact on healthcare utilisation.

Strengths and limitations of this study

To our knowledge, this study was among the first in low-and-middle-income countries (LMICs) that used a national database to examine the impact of COVID-19 infection (as reported in hospital claims data) on healthcare utilisations. Compared to self-reported measure as usually collected in surveys, the use of an administrative database would minimise recall bias.

However, we are also aware that our study has several limitations. First, the study population was limited to NHI members who had at least one regular hospital encounter before being treated with laboratory-confirmed COVID-19. This made our sample smaller than the cumulative number of confirmed cases in Indonesia which were about 1.3 million cases as of February 2021. Second, our comparison group might be imperfect due to unmeasured time-varying factors that differentiate our COVID-19 group from the comparison group; hence residual confounding might still exist. For example, change in income or risk attitude–which was unmeasured in our dataset–might affect healthcare utilisation as well as the timing individuals were infected with COVID-19. We believe an improved dataset that covers all NHI members with more detailed individuals’ characteristics might address some of these issues. Therefore, further cooperation with the Social Security Administrative Body of Health (BPJS-Kesehatan) and the Ministry of Health needs to be initiated. Third, our results around hospital costs are inherently sensitive to the change in clinical practice and the unit cost used for claim payments. Hence, our results should be carefully interpreted beyond the Indonesian context and the examined period. Lastly, due to the lack of information on symptoms and detailed resource utilisation, we could not identify whether patients developed Severe Acute Respiratory Infection (SARI) when treated with laboratory-confirmed COVID-19 infection.

Policy implication

This study provided evidence of a short-term increase in utilisation of hospital services before and after COVID-19 hospitalisation in Indonesia that can be used for future health resource planning. This finding particularly important since the government of Indonesia is planning to shift the budget from the centrally allocated COVID-19 emergency fund to the regular NHI fund once the pandemic ends. As a social health insurance program that combines a subsidised scheme for the poor and a non-subsidised scheme based on a contributory mechanism for the non-poor (either through salary deduction or monthly premiums), it needs to anticipate the total cost of treating COVID-19 patients is required to keep its financial balance. For example, an additional government budget might be needed when a future outbreak arises and the estimated total costs of providing COVID-19 treatment exceed the available NHI fund.

Conclusion

Compared to the comparison group, individuals with COVID-19 that required hospital treatment appeared to have a higher utilisation of outpatient and inpatient services a few months before and after the month of treatment. This finding indicated that people who recently used more healthcare services were in a higher likelihood of contracting with COVID-19, either due to the worse health condition to begin with or because of the increasing contact with infected patients in the healthcare settings. The post-COVID pattern need to be considered by policymakers when estimating the full impact of COVID-19 infection on healthcare resource utilisation, especially in more resource-constrained settings.

Supporting information

S1 Table. Summary of dataset used in the manuscript to create Figs 3 and 4.

(PDF)

pone.0305835.s001.pdf^{(228.8KB, pdf)}

S2 Table. Crude hospital utilisation rate and cost over time.

(PDF)

pone.0305835.s002.pdf^{(122.1KB, pdf)}

Acknowledgments

We thank Social Security Administrative Body of Health (BPJS-Kesehatan) team for data and anonymous reviewers for their constructive feedback and helped to improve this manuscript.

Data Availability

All relevant data used to produce the graphs in the paper are provided in S1 and S2 Tables. The raw data that support the findings of this study are available from Indonesia’s Social Security Administrative Body of Health (BPJS-Kesehatan), which were used under license for the current study, and so are not publicly available. Researchers can request access to the data by contacting BPJS-Kesehatan at ppid@bpjs-kesehatan.go.id. Access to the dataset is subject to approval by BPJS-Kesehatan

Funding Statement

This study is fully funded by Indonesia’s Social Security Administrative Body of Health (BPJS-Kesehatan). Award/grant number 456/BA/0621. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

References

1.Cascella M, Rajnik M, Aleem A, Dulebohn SC, Di Napoli R. Features, Evaluation, and Treatment of Coronavirus (COVID-19). StatPearls. Treasure Island (FL); 2023. [PubMed] [Google Scholar]
2.Li J, Huang DQ, Zou B, Yang H, Hui WZ, Rui F, et al. Epidemiology of COVID-19: A systematic review and meta-analysis of clinical characteristics, risk factors, and outcomes. J Med Virol. 2021;93(3):1449–58. Epub 2020/08/14. doi: 10.1002/jmv.26424 . [DOI] [PMC free article] [PubMed] [Google Scholar]
3.Stokes EK, Zambrano LD, Anderson KN, Marder EP, Raz KM, El Burai Felix S, et al. Coronavirus Disease 2019 Case Surveillance—United States, January 22-May 30, 2020. MMWR Morb Mortal Wkly Rep. 2020;69(24):759–65. Epub 2020/06/20. doi: 10.15585/mmwr.mm6924e2 . [DOI] [PMC free article] [PubMed] [Google Scholar]
4.Surendra H, Elyazar IR, Djaafara BA, Ekawati LL, Saraswati K, Adrian V, et al. Clinical characteristics and mortality associated with COVID-19 in Jakarta, Indonesia: A hospital-based retrospective cohort study. Lancet Reg Health West Pac. 2021;9:100108. Epub 2021/03/09. doi: 10.1016/j.lanwpc.2021.100108 . [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Surendra H, Paramita D, Arista NN, Putri AI, Siregar AA, Puspaningrum E, et al. Geographical variations and district-level factors associated with COVID-19 mortality in Indonesia: a nationwide ecological study. BMC Public Health. 2023;23(1):103. Epub 2023/01/15. doi: 10.1186/s12889-023-15015-0 . [DOI] [PMC free article] [PubMed] [Google Scholar]
6.Taquet M, Dercon Q, Luciano S, Geddes JR, Husain M, Harrison PJ. Incidence, co-occurrence, and evolution of long-COVID features: A 6-month retrospective cohort study of 273,618 survivors of COVID-19. PLoS Med. 2021;18(9):e1003773. Epub 2021/09/29. doi: 10.1371/journal.pmed.1003773 . [DOI] [PMC free article] [PubMed] [Google Scholar]
7.Chen C, Haupert SR, Zimmermann L, Shi X, Fritsche LG, Mukherjee B. Global Prevalence of Post COVID-19 Condition or Long COVID: A Meta-Analysis and Systematic Review. J Infect Dis. 2022. Epub 2022/04/17. doi: 10.1093/infdis/jiac136 . [DOI] [PMC free article] [PubMed] [Google Scholar]
8.Ramzi ZS. Hospital readmissions and post-discharge all-cause mortality in COVID-19 recovered patients; A systematic review and meta-analysis. Am J Emerg Med. 2022;51:267–79. Epub 2021/11/16. doi: 10.1016/j.ajem.2021.10.059 . [DOI] [PMC free article] [PubMed] [Google Scholar]
9.Humphreys H, Kilby L, Kudiersky N, Copeland R. Long COVID and the role of physical activity: a qualitative study. BMJ Open. 2021;11(3):e047632. Epub 2021/03/12. doi: 10.1136/bmjopen-2020-047632 . [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Nandasena H, Pathirathna ML, Atapattu A, Prasanga PTS. Quality of life of COVID 19 patients after discharge: Systematic review. PLoS One. 2022;17(2):e0263941. Epub 2022/02/17. doi: 10.1371/journal.pone.0263941 . [DOI] [PMC free article] [PubMed] [Google Scholar]
11.Nopp S, Moik F, Klok FA, Gattinger D, Petrovic M, Vonbank K, et al. Outpatient Pulmonary Rehabilitation in Patients with Long COVID Improves Exercise Capacity, Functional Status, Dyspnea, Fatigue, and Quality of Life. Respiration. 2022;101(6):593–601. Epub 2022/02/25. doi: 10.1159/000522118 . [DOI] [PMC free article] [PubMed] [Google Scholar]
12.Menges D, Ballouz T, Anagnostopoulos A, Aschmann HE, Domenghino A, Fehr JS, et al. Burden of post-COVID-19 syndrome and implications for healthcare service planning: A population-based cohort study. PLoS One. 2021;16(7):e0254523. Epub 2021/07/13. doi: 10.1371/journal.pone.0254523 . [DOI] [PMC free article] [PubMed] [Google Scholar]
13.Ayoubkhani D, Khunti K, Nafilyan V, Maddox T, Humberstone B, Diamond I, et al. Post-covid syndrome in individuals admitted to hospital with covid-19: retrospective cohort study. BMJ. 2021;372:n693. Epub 2021/04/02. doi: 10.1136/bmj.n693 . [DOI] [PMC free article] [PubMed] [Google Scholar]
14.Magnusson K, Skyrud KD, Suren P, Greve-Isdahl M, Stordal K, Kristoffersen DT, et al. Healthcare use in 700 000 children and adolescents for six months after covid-19: before and after register based cohort study. BMJ. 2022;376:e066809. Epub 2022/01/19. doi: 10.1136/bmj-2021-066809 . [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Skyrud KD, Hernaes KH, Telle KE, Magnusson K. Impacts of mild COVID-19 on elevated use of primary and specialist health care services: A nationwide register study from Norway. PLoS One. 2021;16(10):e0257926. Epub 2021/10/09. doi: 10.1371/journal.pone.0257926 . [DOI] [PMC free article] [PubMed] [Google Scholar]
16.Agustina R, Dartanto T, Sitompul R, Susiloretni KA, Suparmi, Achadi EL, et al. Universal health coverage in Indonesia: concept, progress, and challenges. The Lancet. 2019;393(10166):75–102. doi: 10.1016/s0140-6736(18)31647-7 [DOI] [PubMed] [Google Scholar]
17.Medic S, Anastassopoulou C, Lozanov-Crvenkovic Z, Vukovic V, Dragnic N, Petrovic V, et al. Risk and severity of SARS-CoV-2 reinfections during 2020–2022 in Vojvodina, Serbia: A population-level observational study. Lancet Reg Health Eur. 2022;20:100453. Epub 20220701. doi: 10.1016/j.lanepe.2022.100453 . [DOI] [PMC free article] [PubMed] [Google Scholar]
18.Hendarwan H, Syachroni S, Aryastami NK, Su’udi A, Susilawati MD, Despitasari M, et al. Assessing the COVID-19 diagnostic laboratory capacity in Indonesia in the early phase of the pandemic. WHO South East Asia J Public Health. 2020;9(2):134–40. Epub 2020/09/27. doi: 10.4103/2224-3151.294307 . [DOI] [PubMed] [Google Scholar]
19.Caniglia EC, Murray EJ. Difference-in-Difference in the Time of Cholera: a Gentle Introduction for Epidemiologists. Curr Epidemiol Rep. 2020;7(4):203–11. Epub 2021/04/02. doi: 10.1007/s40471-020-00245-2 . [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Wing C, Simon K, Bello-Gomez RA. Designing Difference in Difference Studies: Best Practices for Public Health Policy Research. Annu Rev Public Health. 2018;39:453–69. Epub 2018/01/13. doi: 10.1146/annurev-publhealth-040617-013507 . [DOI] [PubMed] [Google Scholar]
21.Rothbard S, Etheridge JC, Murray EJ. A Tutorial on Applying the Difference-in-Differences Method to Health Data. Current Epidemiology Reports. 2023. doi: 10.1007/s40471-023-00327-x [DOI] [Google Scholar]
22.Driver JA, Djousse L, Logroscino G, Gaziano JM, Kurth T. Incidence of cardiovascular disease and cancer in advanced age: prospective cohort study. BMJ. 2008;337:a2467. Epub 2008/12/11. doi: 10.1136/bmj.a2467 . [DOI] [PMC free article] [PubMed] [Google Scholar]
23.Lavery AM, Preston LE, Ko JY, Chevinsky JR, DeSisto CL, Pennington AF, et al. Characteristics of hospitalized COVID-19 patients discharged and experiencing same-hospital readmission—United States, March–August 2020. Morbidity and Mortality Weekly Report. 2020;69(45):1695. doi: 10.15585/mmwr.mm6945e2 [DOI] [PMC free article] [PubMed] [Google Scholar]
24.Donnelly JP, Wang XQ, Iwashyna TJ, Prescott HC. Readmission and Death After Initial Hospital Discharge Among Patients With COVID-19 in a Large Multihospital System. JAMA. 2021;325(3):304–6. Epub 2020/12/15. doi: 10.1001/jama.2020.21465 . [DOI] [PMC free article] [PubMed] [Google Scholar]
25.Iannaccone S, Alemanno F, Houdayer E, Brugliera L, Castellazzi P, Cianflone D, et al. COVID-19 rehabilitation units are twice as expensive as regular rehabilitation units. J Rehabil Med. 2020;52(6):jrm00073. Epub 2020/06/10. doi: 10.2340/16501977-2704 . [DOI] [PubMed] [Google Scholar]
26.Lund LC, Hallas J, Nielsen H, Koch A, Mogensen SH, Brun NC, et al. Post-acute effects of SARS-CoV-2 infection in individuals not requiring hospital admission: a Danish population-based cohort study. The Lancet Infectious Diseases. 2021;21(10):1373–82. doi: 10.1016/S1473-3099(21)00211-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
27.Cohen K, Ren S, Heath K, Dasmarinas MC, Jubilo KG, Guo Y, et al. Risk of persistent and new clinical sequelae among adults aged 65 years and older during the post-acute phase of SARS-CoV-2 infection: retrospective cohort study. BMJ. 2022;376:e068414. Epub 2022/02/11. doi: 10.1136/bmj-2021-068414 . [DOI] [PMC free article] [PubMed] [Google Scholar]
28.Gao YD, Ding M, Dong X, Zhang JJ, Kursat Azkur A, Azkur D, et al. Risk factors for severe and critically ill COVID-19 patients: A review. Allergy. 2021;76(2):428–55. Epub 2020/11/14. doi: 10.1111/all.14657 . [DOI] [PubMed] [Google Scholar]
29.van Empel G, Mulyanto J, Wiratama BS. Undertesting of COVID-19 in Indonesia: what has gone wrong? J Glob Health. 2020;10(2):020306. Epub 2020/10/29. doi: 10.7189/jogh.10.020306 . [DOI] [PMC free article] [PubMed] [Google Scholar]
30.Ariawan I, Jusril H. COVID-19 in Indonesia: where are we? Acta Medica Indonesiana. 2020;52(3):193. [PubMed] [Google Scholar]
31.Djalante R, Lassa J, Setiamarga D, Sudjatma A, Indrawan M, Haryanto B, et al. Review and analysis of current responses to COVID-19 in Indonesia: Period of January to March 2020. Prog Disaster Sci. 2020;6:100091. Epub 2020/04/01. doi: 10.1016/j.pdisas.2020.100091 . [DOI] [PMC free article] [PubMed] [Google Scholar]
32.Bellelli G, Rebora P, Valsecchi MG, Bonfanti P, Citerio G, members C-MT. Frailty index predicts poor outcome in COVID-19 patients. Intensive Care Med. 2020;46(8):1634–6. Epub 20200525. doi: 10.1007/s00134-020-06087-2 . [DOI] [PMC free article] [PubMed] [Google Scholar]
33.Nugraha RR, Pratiwi MA, Al-Faizin RE, Permana AB, Setiawan E, Farianty Y, et al. Predicting the cost of COVID-19 treatment and its drivers in Indonesia: analysis of claims data of COVID-19 in 2020–2021. Health Econ Rev. 2022;12(1):45. Epub 2022/09/01. doi: 10.1186/s13561-022-00392-w . [DOI] [PMC free article] [PubMed] [Google Scholar]
34.Ministry of Health. PETUNJUK TEKNIS KLAIM PENGGANTIAN BIAYA PERAWATAN PASIEN PENYAKIT INFEKSI EMERGING TERTENTU BAGI RUMAH SAKIT YANG MENYELENGGARAKAN PELAYANAN CORONAVIRUS DISEASE 2019. Jakarta2020.
35.Dagens A, Sigfrid L, Cai E, Lipworth S, Cheng V, Harris E, et al. Scope, quality, and inclusivity of clinical guidelines produced early in the covid-19 pandemic: rapid review. BMJ. 2020;369:m1936. Epub 2020/05/28. doi: 10.1136/bmj.m1936 . [DOI] [PMC free article] [PubMed] [Google Scholar]
36.Whittaker HR, Gulea C, Koteci A, Kallis C, Morgan AD, Iwundu C, et al. GP consultation rates for sequelae after acute covid-19 in patients managed in the community or hospital in the UK: population based study. BMJ. 2021;375:e065834. Epub 2021/12/31. doi: 10.1136/bmj-2021-065834 . [DOI] [PMC free article] [PubMed] [Google Scholar]
37.Arifin B, Anas T. Lessons learned from COVID-19 vaccination in Indonesia: experiences, challenges, and opportunities. Hum Vaccin Immunother. 2021;17(11):3898–906. Epub 2021/10/07. doi: 10.1080/21645515.2021.1975450 . [DOI] [PMC free article] [PubMed] [Google Scholar]

PLoS One. doi: 10.1371/journal.pone.0305835.r001

Decision Letter 0

Nemer Badwan

22 Mar 2023

PONE-D-22-31470Hospital services utilisation and cost before and after COVID-19 hospital treatment: evidence from IndonesiaPLOS ONE

Dear Dr. Firdaus Hafidz,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

==============================

Please submit your revised manuscript by May 06 2023 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

With Kind Regards,

Asst. Prof. Dr. Nemer Badwan

PhD in Economics and Finance

Assistant Professor of Economics and Finance

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Please provide additional details regarding participant consent. In the ethics statement in the Methods and online submission information, please ensure that you have specified (1) whether consent was informed and (2) what type you obtained (for instance, written or verbal, and if verbal, how it was documented and witnessed). If your study included minors, state whether you obtained consent from parents or guardians. If the need for consent was waived by the ethics committee, please include this information.

If you are reporting a retrospective study of medical records or archived samples, please ensure that you have discussed whether all data were fully anonymized before you accessed them and/or whether the IRB or ethics committee waived the requirement for informed consent. If patients provided informed written consent to have data from their medical records used in research, please include this information.

3. Thank you for stating the following financial disclosure:

"This study is fully funded by Indonesia’s Social Security Administrative Body of Health (BPJS-Kesehatan). Award/grant number 456/BA/0621"

Please state what role the funders took in the study. If the funders had no role, please state: "The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript."

If this statement is not correct you must amend it as needed.

Please include this amended Role of Funder statement in your cover letter; we will change the online submission form on your behalf.

4. Your ethics statement should only appear in the Methods section of your manuscript. If your ethics statement is written in any section besides the Methods, please move it to the Methods section and delete it from any other section. Please ensure that your ethics statement is included in your manuscript, as the ethics statement entered into the online submission form will not be published alongside your manuscript.

5. In your Data Availability statement, you have not specified where the minimal data set underlying the results described in your manuscript can be found. PLOS defines a study's minimal data set as the underlying data used to reach the conclusions drawn in the manuscript and any additional data required to replicate the reported study findings in their entirety. All PLOS journals require that the minimal data set be made fully available. For more information about our data policy, please see http://journals.plos.org/plosone/s/data-availability.

Upon re-submitting your revised manuscript, please upload your study’s minimal underlying data set as either Supporting Information files or to a stable, public repository and include the relevant URLs, DOIs, or accession numbers within your revised cover letter. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories. Any potentially identifying patient information must be fully anonymized.

Important: If there are ethical or legal restrictions to sharing your data publicly, please explain these restrictions in detail. Please see our guidelines for more information on what we consider unacceptable restrictions to publicly sharing data: http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions. Note that it is not acceptable for the authors to be the sole named individuals responsible for ensuring data access.

We will update your Data Availability statement to reflect the information you provide in your cover letter.

6. Please upload a new copy of Figure xxxx as the detail is not clear. Please follow the link for more information: https://blogs.plos.org/plos/2019/06/looking-good-tips-for-creating-your-plos-figures-graphics/ " https://blogs.plos.org/plos/2019/06/looking-good-tips-for-creating-your-plos-figures-graphics/.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: No

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: This paper presents important information on hospital services utilization and cost before and after COVID-19 hospital treatment. The article is well written but there is some minor editing required. Also, I have highlighted a few areas of concerns in the submitted manuscript (see uploaded attachment.

Reviewer #2: This is an informative and well written article. There are a number of suggestions for strengthening the content and findings.

1. In the introduction and discussion I would like to see more information on the population based epidemiology of covid-19 for Indonesia. This will help to better understand the landscape at this time. This study occurred during the first surges globally and the incidence and prevalence of the condition and other demographic characteristics of those effected and uneffected would be useful to see.

2. The data is based upon a large administrative data base from Indonesia subjects included "with 28,159 Indonesian NHI enrollees treated with laboratory-confirmed COVID-19, compared with 8,995 individuals never diagnosed with COVID-19 in 2020." The methods used for laboratory confirmed Covid-19 dx is not described in detail. Was this according to the WHO criteria for diagnosing Covid-19. In addition, as this was based upon national data, were the laboratory methods across the country uniform and standardized?

3. More detail is needed for understanding the Covid-19 positive and "never diagnosed" groups. What is the level of comparability for the two groups (covid positive and "never diagnosed" beingcompared). Perhaps a propensity analysis would be useful to better understand the distance for all of the sociodemographic variables included to best understand the differences between the two groups compared. When matching I would base this on the Mahalanobis distance (MD), with the selection of the caliper. The study is only as good as an indepth understanding of the control group.

4. For the control group, what is meant by "never diagnosed." Does this mean that those in the control group received the Covid-19 test and were negative? Is it possible that there was a fraction of individuals in the control group that were positive for covid?

4. An instrumental variable analysis might be useful to better understand unmeasured variables that go uncontrolled for in the analysis? What might such unmeasured variables be?

5. As this study was done during the earlier surges of the global pandemic it would be useful to understand the severity of the presenting covid-19 and if data is available have a breakdown of costs and hospitalizations for those severely impacted.

6. The focus for stratifications for hospital costs of services is by age (< 40 and >= 40). How was 40 years of age selected as the cutoff? Could more stratifications be included for age by a younger age group and the elderly? Results may be interesting. Can additional stratifications for other variables be included such as gender and educational level (a proxy for income).

7. I would like to see the Indonesian currency "Rupiahs" converted to US dollars. Or at least a footnote indicating the conversion rate.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Adebola Emmanuel Orimadegun

Reviewer #2: No

**********

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

<quillbot-extension-portal></quillbot-extension-portal>

Attachment

Submitted filename: PONE-D-22-31470_reviewer.pdf

pone.0305835.s003.pdf^{(1.3MB, pdf)}

PLoS One. 2024 Jul 5;19(7):e0305835. doi: 10.1371/journal.pone.0305835.r002

Author response to Decision Letter 0

20 May 2023

Dear Editor,

Thank you for considering our article for publication at PLOS One. We really appreciate the comments and suggestions from you and the reviewers.

Editor

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf

and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

Thank you for the reminder. We have read the requirements and edited our files accordingly.

We have added in Methods section Ethical Clearance with the sentence as follows,

“All data were fully anonymized before being accessed.”

3. Thank you for stating the following financial disclosure:

"This study is fully funded by Indonesia’s Social Security Administrative Body of Health (BPJS-Kesehatan). Award/grant number 456/BA/0621"

If this statement is not correct you must amend it as needed.

Please include this amended Role of Funder statement in your cover letter; we will change the online submission form on your behalf.

We have put the additional information in the funding section as mentioned in the cover letter

"The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.”

We have moved the ethics statement to the Methods section.

We have updated our data availability statement to “All relevant data are within the manuscript and its Supporting Information files.” The relevant data are now submitted as supporting file.

we have reuploaded our figures to follow the requirements with the maximum quality we could produce.

Reviewer 1

This paper presents important information on hospital services utilization and cost before and after COVID-19 hospital treatment. The article is well written but there is some minor editing required. Also, I have highlighted a few areas of concerns in the submitted manuscript (see uploaded attachment.

Thank you for the comments and suggestions. We have addressed all of them in the revised manuscript. Here the detail:

1. Introduction, Paragraph 1:

- We have replaced the sentence “A study in the US also mentioned that COVID-19 hospitalisation among patients with preexisting comorbidities was six times higher compared to those without ones” with a new sentence “A study conducted in Unites Stated found that COVID-19 hospitalization was six times higher in patients with preexisting comorbidities than in those without”

- We edited the last sentence and delete the sentence of “The same pattern also emerged from multiple settings, including in low-and-middle-income countries (LMICs),”

2. Introduction, paragraph 2:

- We have deleted the word “also” in the first sentence

- We edited the sentence “Due to its multi-organ involvement, some COVID-19 symptoms can persist up to six months after infection remission or more, defined as long covid or post-covid syndrome” with a new sentence, “Due to its multi-organ involvement, some COVID-19 symptoms can persist up to six months after infection remission or more, which is defined as long covid or post-covid syndrome”

3. Methods, Section Study Population

We have added the comma after the word care in the sentence as follows

“Individuals who were diagnosed based on RT-PCR positive results and treated, either in outpatient care or admitted to inpatient care, from May to August 2020 were included in the COVID-19 group.”

4. Figure 1’s Note

We have moved the note to the Study Period’s paragraph as follows,

“We considered a period of August 2019 to November 2020 in our analysis. December 2020 and January 2021 were excluded due to the possibility of elevated healthcare utilisation two months before individuals in the control group were hospitalised with COVID-19 in February 2021. Hence, we had a maximum of six months (June 2020 to November 2020) and a minimum of three months (September to November 2020) post-COVID-19 follow-up. For pre-COVID period, we included nine months before COVID-19 hospital treatment. The grouping strategy and the period of observation used in our analyses is shown in Fig 1”

5. Figure 2’s Note

We have moved the note to the Study Participants’ paragraph as follows,

“Figure 2 depicts the flow of study participants and their classification into COVID-19 and control group. “

6. Result, Section Changes in Utilisation of Hospital Services

We have deleted the sentence “The fifth month leading to the COVID-19 event was used as the reference period. We assumed the difference in outcome between the case and control in other periods to be constant” and put this explanation in the Methods section Study Population as follows:

“Those who treated in hospitals with COVID-19 in February 2021 were classified as the control group, assuming they never contracted COVID-19 in 2020. Note that in the NHI database, only results from RT-PCR (Reverse Transcriptase Polymerase Chain Reaction) done by Ministry of Health-accredited laboratories are used.”

7. Results, Section Changes in Total Cost of Hospital Services

We have deleted the sentence, “In our regression, we used the period of 6 to 4 months before COVID-19 hospitalisation as the reference time where we assume no difference in admissions rate between the COVID-19 and the control group. We also lumped together 9 to 7 months before COVID19 to see whether there was any gap in outcomes between the two groups. “

The explanation related to the sentence has been included in the Methods, Section Study Period:

8. Discussion, Section Main Findings

- We have deleted this sentence, “Among 28,155 Indonesian National Health Insurance (NHI) enrollees who were in hospitals with laboratory-confirmed COVID-19 between May and August 2020”

- We revised the sentence “Specifically, we found an increase outpatient utilisation that happened three months before and up to three months after the treatment and an uptick in admissions rate one month before and after the treatment” with the additional word suggested by the reviewer as follows,

“Specifically, we found an increase in outpatient utilisation that happened three months before and up to three months after the treatment and an uptick in admissions rate one month before and after the treatment.”

- For the sentences related to the literature review, we have added some explanatory sentences to contrast their findings with ours, as follows

“Other studies reported that 20% of COVID-19 patients might require rehospitalisation, especially in older individuals and those with comorbidities [21-23]. This finding was consistent with our results where higher post-COVID utilisation was observed mainly in the older patients. In contrast, two studies in Norway and Denmark that only included mild cases did not find a short-term increase in specialist care [14, 15, 24].

9. Discussion, Section Policy Implication

We have replaced the word “elevation” with “to increase”

10. Conclusion

We have revised our conclusion not to recite our results. Therefore, our conclusion has been written as follows,

“Compared to the control group, individuals with COVID-19 that required hospital treatment appeared to have a higher utilisation of outpatient and inpatient services a few months before and after the month of treatment. This finding indicated that people who recently used more healthcare services were in a higher likelihood of contracting with COVID, either due to the worse health condition to begin with or because of the increasing contact with infected patients in the healthcare settings. More research is needed to fully understand the possibility of predicting COVID-19 infection based on prior healthcare utilisation. Lastly, the post-COVID pattern also need to be considered by policymakers when estimating the full impact of COVID-19 infection on healthcare resource utilisation, especially in more resource-constrained settings.”

Reviewer 2

This is an informative and well written article. There are a number of suggestions for strengthening the content and findings.

Thank you for your comments and suggestions. We have addressed all of them and provide details of our responses below.

1. In the introduction and discussion, I would like to see more information on the population based epidemiology of covid-19 for Indonesia. This will help to better understand the landscape at this time. This study occurred during the first surges globally and the incidence and prevalence of the condition and other demographic characteristics of those effected and unaffected would be useful to see.

In the introduction, we have added the following sentence to briefly lay out the epidemiology of COVID-19 in Indonesia, particularly in the earlier period.

“In Indonesia, about 12% of the hospitalised patients died in the first three months of the pandemic, with higher rate among older people, male, and those with multiple comorbidities [4]. A more recent study indicated that districts with higher proportion of elderly and lower healthcare capacity had higher rate of COVID-19 mortality [5].”

We have clarified this issue in our revised manuscript by adding the following sentence in our method section.

“Note that in the NHI database, only results from RT-PCR (Reverse Transcriptase Polymerase Chain Reaction) done by Ministry of Health-accredited laboratories are used.”

3. More detail is needed for understanding the Covid-19 positive and "never diagnosed" groups. What is the level of comparability for the two groups (covid positive and "never diagnosed" being compared). Perhaps a propensity analysis would be useful to better understand the distance for all of the sociodemographic variables included to best understand the differences between the two groups compared. When matching I would base this on the Mahalanobis distance (MD), with the selection of the caliper. The study is only as good as an indepth understanding of the control group.

We agree that the comparability in characteristics between the Covid-19 group and the “never diagnosed” or control group is important. We have shown in Table 1 (and explained in the Result section) that there are several characteristics differences. However, we believe that we have addressed these observed variations by including them as control variables in our regressions. We further highlight our approach by revising our statistical analysis subsection.

In our results, we have also shown that parallel pre-trends – which is one way to indicate that our DiD was valid – were mostly satisfied. Therefore, we decide not to proceed with matching on characteristics before employing the DiD estimation. For example, we have now mentioned

“Our results indicated that up to the reference month, the differences in outcome between the COVID-19 and control group were close to zero and all statistically insignificant. This finding supported our assumption that the trend in outcome between the two groups were similar in the absence of COVID-19.”

The control group was the people that were diagnosed in February 2021, they

Attachment

Submitted filename: Responses to Reviewers.docx

pone.0305835.s004.docx^{(46.4KB, docx)}

PLoS One. doi: 10.1371/journal.pone.0305835.r003

Decision Letter 1

Nemer Badwan

1 Sep 2023

PONE-D-22-31470R1Hospital services utilisation and cost before and after COVID-19 hospital treatment: evidence from IndonesiaPLOS ONE

Dear Dr. Firdaus Hafidz,

Please submit your revised manuscript by Oct 16 2023 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

We look forward to receiving your revised manuscript.

Kind Regards,

Asst. Prof. Dr. Nemer Badwan

PhD in Economics and Finance

Assistant Professor of Economics and Finance

Academic Editor

PLOS ONE

Journal Requirements:

Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript. If you need to cite a retracted article, indicate the article’s retracted status in the References list and also include a citation and full reference for the retraction notice.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

Reviewer #3: (No Response)

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

Reviewer #1: Yes

Reviewer #3: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #3: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

Reviewer #1: Yes

Reviewer #3: No

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

Reviewer #1: Yes

Reviewer #3: Yes

**********

6. Review Comments to the Author

Reviewer #1: (No Response)

Reviewer #3: The authors conducted a retrospective study on parameters related to COVID-19 healthcare in Indonesia. The idea to focus on periods before or after treatment of COVID-19 patients is highly interesting. Scope and idea of the study are clearly described. The study is relevant for understanding the pandemic’s impact of the pandemic on the healthcare system.

Comments/questions:

1. I suppose “COVID-19” here is defined as a lab-confirmed SARS-CoV-2 infection. Are information available that allow for identifying patients who actually developed corresponding respiratory systems? What might be the influence of patients without SARI?

2. If “COVID-19” were indeed defined as SARS-CoV-2 infection, I would recommend including SARI as a comorbidity too.

3. How does hospital service utilization enter the statistical models? Is it a binary variable or a count per month? What kind of model was used, logistic or Poisson regression? Please add more details.

4. Is a linear model used for total claims (costs)? Did you check the distribution of costs? Was the variable transformed before entering the model? In my experience, this variable is highly skewed. Please add more details.

5. The model formula (line 146) contains two coefficients, alpha and beta. I suppose the regression estimated coefficients for the other covariates too. If this is true, please extend the formula.

6. The model formula appears to contain interaction terms between the month and several covariates (age, sex, comorbidities). Did the model also contain terms for main effects of these covariates?

7. I would like to know whether the authors did conduct the analyses stratified for total, age groups, and sex.

8. The authors present the proportions of different comorbidities. Are these comorbidities derived from ICD codes?

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Adebola E. Orimadegun

Reviewer #3: No

**********

PLoS One. 2024 Jul 5;19(7):e0305835. doi: 10.1371/journal.pone.0305835.r004

Author response to Decision Letter 1

15 Oct 2023

Dear Asst. Prof. Dr. Nemer Badwan,

I am writing to express my gratitude for the second valuable feedback provided by the reviewers and yourself regarding our submitted manuscript titled "Hospital services utilisation and cost before and after COVID-19 hospital treatment: evidence from Indonesia." Here below our details comments and response:

Journal Requirements

We have checked all references, and found no retracted references.

Reviewers' comments

Reviewer's Responses to Questions

Comments to the Author

Reviewer #1: All comments have been addressed.

Reviewer #3: (No Response)

2. Is the manuscript technically sound, and do the data support the conclusions?

Reviewer #1: Yes

Reviewer #3: Yes

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #3: Yes

4. Have the authors made all data underlying the findings in their manuscript fully available?

Reviewer #1: Yes

Reviewer #3: No

We have revised our data sharing statement to the following.

All relevant data used to produce the graphs in the paper are provided in S1 Table. The raw data that support the findings of this study are available from by Indonesia’s Social Security Administrative Body of Health (BPJS-Kesehatan), which were used under license for the current study, and so are not publicly available. Data is available from: https://data.bpjs-kesehatan.go.id/ by applying via the website. Other researchers will be able to access the data set in the same way as the authors.

5. Is the manuscript presented in an intelligible fashion and written in standard English?

Reviewer #1: Yes

Reviewer #3: Yes

6. Review Comments to the Author

Reviewer #1: (No Response)

Reviewer #3:

Reviewer 3

The authors conducted a retrospective study on parameters related to COVID-19 healthcare in Indonesia. The idea to focus on periods before or after treatment of COVID-19 patients is highly interesting. Scope and idea of the study are clearly described. The study is relevant for understanding the pandemic’s impact of the pandemic on the healthcare system.

We thank the reviewer for the valuable comments. We have addressed each of your concerns below.

I suppose “COVID-19” here is defined as a lab-confirmed SARS-CoV-2 infection. Are information available that allow for identifying patients who actually developed corresponding respiratory systems? What might be the influence of patients without SARI?

Thank you for raising the issue that there might be some laboratory-confirmed SARS-CoV-2 infections that did not develop into Severe Acute Respiratory Infection (SARI). Unfortunately, we do not have information on symptoms or detailed treatment received by the patients to identify SARI. Therefore, classifying COVID-19 cases with and without SARI is not possible. We have added this limitation to our revised manuscript.

“Lastly, due to the lack of information on symptoms and detailed resource utilisation, we could not identify whether patients developed Severe Acute Respiratory Infection (SARI) when treated with laboratory-confirmed COVID-19 infection.”

If “COVID-19” were indeed defined as SARS-CoV-2 infection, I would recommend including SARI as a comorbidity too.

Because we cannot identify SARI, we cannot include it as comorbidity in our analysis. Similar to your first point, we have added this limitation to our revised manuscript.

How does hospital service utilization enter the statistical models? Is it a binary variable or a count per month? What kind of model was used, logistic or Poisson regression? Please add more details.

Thank you for pointing out this issue. We are aware that the hospital service utilisation is a count data, representing the number of visits per individual per month. We have added the following sentence in our method section to clarify this concern:

“Although some of our outcomes were count variables, linear modelling can still be an appropriate choice, hence we estimate Equation 1 using Ordinary Least Square (OLS).”

Rothbard, S., Etheridge, J.C. & Murray, E.J. A Tutorial on Applying the Difference-in-Differences Method to Health Data. Curr Epidemiol Rep (2023). https://doi.org/10.1007/s40471-023-00327-x

Is a linear model used for total claims (costs)? Did you check the distribution of costs? Was the variable transformed before entering the model? In my experience, this variable is highly skewed. Please add more details.

We agree that the skewness of claims (costs) data exist. However, since our aims is to estimate the quantity in absolute terms (Indonesian Rupiahs), we do not transform the costs variable first. To address this concern, we have added the following sentence to our method section.

“In addition, for costs outcome, we conducted additional analysis where the variable is first transformed into a log form because of the skewness of the data.”

We then described the findings in our result section.

“These results were similar when we used log-transformed value of the costs variables.”

The model formula (line 146) contains two coefficients, alpha and beta. I suppose the regression estimated coefficients for the other covariates too. If this is true, please extend the formula.

We have extended the regression equation to clarify that we also estimated the coefficients of other covariates. The revised equation is written below:

Y_it=α_i+Σ_(t=Min)^(t=Max) 〖β_1〗_t Month_t×Covid_i+β_2 Covid_i+β_3 Month_t+〖β_4 X〗_i Month_t+〖β_5 X〗_i+ε_it

The model formula appears to contain interaction terms between the month and several covariates (age, sex, comorbidities). Did the model also contain terms for main effects of these covariates?

Yes, it did. We have now clarified the confusion by expanding our regression equation, as we have mentioned in the point number 5. In this new equation, the coefficient for the main effects of the covariates were represented by β_5 vector. However, because these variables were used as control variables, we did not report and discuss the estimated coefficients further.

I would like to know whether the authors did conduct the analyses stratified for total, age groups, and sex.

Yes, we did. We have already mentioned in the section “Statistical Analysis” of our initial manuscript as the following:

“Heterogeneity analysis was conducted by splitting the sample into younger (<40 years old) and older (≥40) population, as well as by gender.”

We have also reported the results for total sample, splitting by age group as well as gender or sex.

The authors present the proportions of different comorbidities. Are these comorbidities derived from ICD codes?

ICD-10 codes of the primary and secondary diagnoses were used to define the comorbidities (see the list below)

Comorbidity ICD-10 Codes

Hypertension I10 – I16

Type 2 Diabetes Mellitus E11

Heart disease I11, I125

Tuberculosis A15 to A19

Asthma J45

COPD J44

Cancer C0 – C9; D1 – D4

Liver disease K70 – K77

Chronic Kidney Disease I12, I13, and N18

Attachment

Submitted filename: Response to Reviewer Sept.docx

pone.0305835.s005.docx^{(24.2KB, docx)}

PLoS One. doi: 10.1371/journal.pone.0305835.r005

Decision Letter 2

Nemer Badwan

17 Oct 2023

PONE-D-22-31470R2Hospital services utilisation and cost before and after COVID-19 hospital treatment: evidence from IndonesiaPLOS ONE

Dear Dr. Hafidz,

==============================

Please submit your revised manuscript by Dec 01 2023 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

We look forward to receiving your revised manuscript.

Kind regards,

Asst. Prof. Dr. Nemer Badwan, Ph.D in Economics and Finance

Academic Editor

PLOS ONE

Journal Requirements:

[Note: HTML markup is below. Please do not edit.]

PLoS One. 2024 Jul 5;19(7):e0305835. doi: 10.1371/journal.pone.0305835.r006

Author response to Decision Letter 2

17 Nov 2023

Dear Reviewer,

Thank you for your invaluable feedback on our second submission. We have carefully revised the manuscript in accordance with your suggestions. We noticed that we did not receive specific feedback on our third response; therefore, we have resubmitted it as it was. Regarding the references, we have thoroughly reviewed them and confirmed that none of the articles have been retracted. If there are any discrepancies in the reference list, please do let us know so we can make the necessary corrections.

Best regards,

Hafidz

Attachment

Submitted filename: Response to Reviewer Sept.docx

pone.0305835.s006.docx^{(24.2KB, docx)}

PLoS One. doi: 10.1371/journal.pone.0305835.r007

Decision Letter 3

Nemer Badwan

6 Dec 2023

PONE-D-22-31470R3Hospital services utilisation and cost before and after COVID-19 hospital treatment: evidence from IndonesiaPLOS ONE

Dear Dr. Hafidz,

==============================

Please submit your revised manuscript by Jan 20 2024 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
A marked-up copy of your manuscript that highlights changes made to the original version You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
An unmarked version of your revised paper without tracked changes should be uploaded as a separate file labeled 'Manuscript'.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions, see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind Regards,

Asst. Prof. Dr. Nemer Badwan

Ph.D in Economics and Finance

Assistant Professor of Economics and Finance

Academic Editor

PLOS ONE

[Note: HTML markup is below.] [Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

Reviewer #5: (No Response)

Reviewer #6: (No Response)

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #5: Partly

Reviewer #6: Partly

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #5: Yes

Reviewer #6: Yes

**********

4. Have the authors made all the data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with a rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g., participant privacy or use of data from a third party—those must be specified.

Reviewer #5: Yes

Reviewer #6: No

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

Reviewer #5: Yes

Reviewer #6: Yes

**********

6. Review Comments to the Author

Reviewer #5: Major comments:

- Choice of reference period? What is the rationale for using the fifth month for hospital utilization and the 4-6 months before for cost? If you had chosen one month prior to hospitalization, the results would have been different. Why is the period in February used as comparison group?

- Is the aim of the study to investigate the utilization and cost for the ‘patients with COVID’ or the ‘COVID-19 disease’ itself? See the following text from your paper: “ We explored the short- to medium-term change in the pre- and post-acute period of COVID-19 treatment to fully understand the extent to which COVID-19 might be associated with the use of hospital services, especially in a resource-constrained setting.”

- I would like to include the crude rate of hospital utilization and cost, not only the estimated difference in the difference plot, to be sure of the assumption that the trend in outcome between the two groups was similar in the absence of COVID-19.

- I also want you to include the figures on the difference in the estimate for cost analysis and not only for health care utilization to see if the pre-trend assumption also holds here.

- In Figure 4 (in-patient, May period), I think the estimates of the difference are not zero in the months prior to the fifth month before COVID-19 treatment but rather increasing; this may be discussed in the Discussion section.

- Additionally, also you should also discuss what causes the increased use of health care and cost from 5 months before treatment; it is a long period, are these frail patients?

- And why are the patients older in the comparison group?

- There are more people in the comparison group with the non-subsidized membership scheme; could this affect health care utilization?

Minor comments:

- Please use the comparison group instead of the control group since it is not an RTC.

- Fig 1, misspelled in legend yellow, Months before COVID-10 hsopitalization

- Severity level: how are the different levels defined? Please include this information in the method section and in the table notation. Is this the same as the COVID-19 severity used as the control variable in the regression? And have all included patients with COVID-19 as their main cause of hospitalization, or could it also be that they were arbitrary was tested when admitted to the hospital?

- In figure 1, it is stated that "lines represented the estimated difference between the COVID-19 and 215 control groups, controlling for demographic characteristics," but the controlling variable is the same as stated in the tables: “Regressions control for individuals’ year of birth, gender, COVID-19 severity, NHI membership segment, comorbidities prior to 2020 interacted with the month indicator, as well as month and province fixed effects.” But in the data source section, the basic demographic characteristics included year of birth, sex, district and province, and type of NHI membership, but nothing about COVID-19 severity or comorbidity. Please indicate all controlling variables in the figure legends.

- Please be consistent with the word, like for instance, province or district instead of region, and COVID-severity instead of only severity.

- Use log-transformed values of the cost variables. What are the results from the additional analysis? I can’t see that Table 2 includes these results. “ These results were similar when we used the log-transformed value of the cost variables (see Table 2).”

Reviewer #6: The study authors conducted an analysis comparing inpatient hospital visit rates, outpatient hospital visit rates, and hospital-associated costs associated with COVID-19-positive patients compared with controls who have not yet contracted COVID-19. I believe the study is methodologically sound, but this can be strengthened with more details regarding data sources and the statistical methods.

- Page 3 line 62: “with” instead of “which” or some rewording of this first sentence is needed

- Page 6 line 128: “in in”

- Page 9 line 184: range of age minimum is 44.2, not 44.6, if I’m reading this correctly

- It took me a while, but after studying Figure 1 and looking at the results, I believe I understand the methods. Please confirm if I’m understanding this correctly. If I am, then I think the methods can be written differently to more simply explain what was done and why

o The control group (unexposed) contracted COVID-19 eventually. This fact is the reason why this group was chosen as a control (as a means to adjust for some confounders)

o The case group (exposed) contracted COVID-19 earlier (2020)

o The difference between exposed and unexposed was determined by time period (-9, -8, 0, 1, …, 6)

- We would expect, on average, that the cases and controls are similar with respect to demographic and clinical characteristics. However, this is generally not true as COVID-19 changed over time (e.g., virulence), immunity changed, and public health mandates changed as well. Standardized differences are not reported but are probably meaningfully different between cases and controls during the “pre” period. DiD methods often use propensity score methods to adjust for such factors. Why was this not considered? This is particularly of concern when, using age as an example

- It appears that there is a lack of parallel trends for the inpatient analysis (Figure 4).

- It’s unclear how clustering was taken into account since the same patient can be found in the analysis multiple times. From the table captions, it seems like there was some adjustment, but I did not see this mentioned in the methods.

- An assumption was made on page 6 that “those who were treated in hospitals with COVID-19 in February 2021 were classified as the control group, assuming they never contracted COVID-19 in 2020," but it’s not clear why an assumption had to be made. Rather, couldn’t this have been verified from the data?

- If coefficient 0.1t represents the month-specific DiD, is this what’s reported in Table 2? It isn’t clear from the table descriptor or the description of the results in the prose. For example, the paragraph beginning on page 17, line 270, does not mention interaction in the interpretation.

- It is not specified in the methods by which cases of COVID-19 were captured or the accuracy of this. Different countries may have had differential access to PCR testing over time, and inpatient testing may be more accurately documented than outpatient testing.

- The descriptors of cost were not presented. What types of costs were captured?

- I don’t think using prior medical history to predict COVID-19 infection should be mentioned in the conclusions.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous, but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #5: No

Reviewer #6: No

**********

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". [If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that supporting information files do not need this step.

PLoS One. 2024 Jul 5;19(7):e0305835. doi: 10.1371/journal.pone.0305835.r008

Author response to Decision Letter 3

11 Feb 2024

Dear Asst. Prof. Dr. Nemer Badwan,

I am writing to express my gratitude for the third valuable feedback provided by the reviewers and yourself regarding our submitted manuscript titled "Hospital services utilisation and cost before and after COVID-19 hospital treatment: evidence from Indonesia." Here below our details comments and response:

Reviewers' comments

Reviewer #5

Major comments:

This is an arbitrary decision. The point of reference period is to show that prior to the reference period, the differences in changes in outcomes over time is close to zero (the parallel pre-trend). This is a matter of presentation. Moreover, referencing Skyrud et al. (2021), no long-term elevation in healthcare use was observed in this period, supporting our timeframe choice for a balanced analysis.

February 2021 was selected as the control group to establish a COVID-19 baseline, excluding December 2020 and January 2021 due to potential elevated healthcare utilization before the comparison group were hospitalized.

It is for patients with COVID especially those who are treated in hospital, either inpatient or outpatient, due to COVID-9 which was confirmed based on standardised laboratory test. To clarify the aim, we revised the sentence:

"Our study aims to examine the impact of COVID-19 by analysing both the direct effects on patients treated for the disease and the associated changes in hospital service utilization and costs. This dual perspective provides insights into the short- to medium-term adjustments within hospital services pre- and post-acute COVID-19 treatment periods, especially critical for understanding the pandemic's ramifications in resource-constrained settings."

Thank you for your suggestion. We have added Tables showing the crude rate of hospital utilisation and cost as supporting information in Table S2.

- I also want you to include the figures on the difference in the estimate for cost analysis and not only for health care utilization to see if the pre-trend assumption also holds here.

Thank you for the suggestion. However, we believe this is unnecessary as our attempt to show the parallel pre-trend has been shown in Tables 2 to 4. Specifically, we have compared the total cost of COVID-19 and the comparison group at the reference period of 4 to 6 months before hospital treatment with a period of 7 to 9 months before. Hence, the estimates for the pre-reference period (7 to 9 months before hospital treatment), shown in the first row of the table, have indicated parallel pre-trends.

As discussed previously, the choice of reference group is a matter of presentation. We believe this does not require further discussion.

- Additionally, also you should also discuss what causes the increased use of health care and cost from 5 months before treatment; it is a long period, are these frail patients?

The increased use of healthcare and costs from 5 months before treatment could suggest that patients with pre-existing conditions are more likely to experience severe COVID-19 outcomes, necessitating more extensive healthcare utilization. This period might capture healthcare interactions related to managing chronic conditions, which are exacerbated or become more complex in the lead-up to a COVID-19 diagnosis. The notion of frail patients being more susceptible to higher healthcare use and costs aligns with broader medical literature, indicating that individuals with compromised health status face greater risks and healthcare needs during infectious disease outbreaks. We added in the discussion part

- And why are the patients older in the comparison group?

We are not sure about this, has not found any evidence support. One possible explanation for the older age profile in the comparison group (those affected in 2021) could be related to the evolution of the pandemic and changes in virus transmission dynamics over time. Initially, efforts might have focused on protecting and isolating older individuals, potentially leading to lower infection rates among this group in 2020. As the pandemic progressed into 2021, with the emergence of new variants or changes in public health measures, older individuals may have become more susceptible or exposed, leading to a higher representation in the comparison group.

- There are more people in the comparison group with the non-subsidized membership scheme; could this affect health care utilization?

Indeed, the presence of more individuals under the non-subsidized membership scheme in the comparison group could impact healthcare utilization patterns. However, it's important to note that during the pandemic, the government covered all patients' treatment costs for COVID-19, regardless of whether they had JKN (Indonesia's National Health Insurance) or not. This policy likely influenced a surge in access to COVID-19 services, with many claims categorized under the non-subsidized scheme. We anticipate this unique situation may have mitigated potential disparities in healthcare access between different insurance membership categories during the pandemic

Minor comments:

- Please use the comparison group instead of the control group since it is not an RTC.

Thanks, we replaced all control group to comparison group

- Fig 1, misspelled in legend yellow, Months before COVID-10 hospitalisation

Thanks, we revised the misspelling.

Severity levels in our study are defined based on the Indonesian Case-Based Groups (INA-CBGs) system, a case-mix payment system that utilizes a software grouper application. The severity level is influenced by complications and comorbidities, which are indicative of the resource intensity level required for treatment during the first treatment episode where patients were laboratory-confirmed. Yes, these severity levels are used as control variables in our regression analyses. All included patients had COVID-19 as their primary cause of hospitalization; the system ensures that the diagnosis is not arbitrary but based on rigorous clinical assessment upon admission.

We included the information in the method section and we had it in table notation

We included in the data source section.

- Please be consistent with the word, like for instance, province or district instead of region, and COVID-severity instead of only severity.

Thanks, we revised the region into province

Apologies for any confusion caused regarding the log-transformation of cost variables. To clarify, the log-transformation of cost data was performed as part of our preliminary analysis to address data skewness, rather than as an additional, separate analysis. However, after tested it has similar results and easier to interpreted in non-log-transformed, ensuring a more accurate and meaningful statistical analysis. We revised the method section and deleted the results sentences.

Reviewer #6

The study authors conducted an analysis comparing inpatient hospital visit rates, outpatient hospital visit rates, and hospital-associated costs associated with COVID-19-positive patients compared with controls who have not yet contracted COVID-19. I believe the study is methodologically sound, but this can be strengthened with more details regarding data sources and the statistical methods.

- Page 3 line 62: “with” instead of “which” or some rewording of this first sentence is needed

Replaced. Thanks,

- Page 6 line 128: “in in”

Replaced. Thanks,

- Page 9 line 184: range of age minimum is 44.2, not 44.6, if I’m reading this correctly

Replaced. Thanks.

o The control group (unexposed) contracted COVID-19 eventually. This fact is the reason why this group was chosen as a control (as a means to adjust for some confounders)

o The case group (exposed) contracted COVID-19 earlier (2020)

o The difference between exposed and unexposed was determined by time period (-9, -8, 0, 1, …, 6)

Thanks for the suggestions. We revised in method section make it simpler way to explain.

We appreciate the reviewer's suggestion to employ propensity score methods (PSM) to adjust for differences between cases and controls. In anticipation of these concerns, we explored the application of PSM in our analysis. However, upon implementation, we encountered significant challenges: the confidence intervals of our estimates became notably wider, reducing the precision of our results. We think it is because the matching process necessitated a reduced sample size, potentially limiting the statistical power and generalizability of our findings. Moreover, we had already accounted for individual characteristics as control variables in our regression model, interacting these with the month variable to robustly address temporal variations in their effects on the outcomes.

- It appears that there is a lack of parallel trends for the inpatient analysis (Figure 4).

Yes correct, we explained in the results section.

Yes, it is true that the same patient may appear multiple times in our analysis due to the monthly clustering approach we adopted. As noted in our Tables, standard error were clustered at individual level.

The assumption that individuals treated for COVID-19 in hospitals in February 2021 had not contracted the virus in 2020 was made to ensure that this is the right comparison group. While theoretically possible to verify past infections, this information is not available in our datasets.

Not sure about the coefficient 0.1t you mentioned, but it the coefficients represented the estimated differences in total costs of hospital services (in thousand Rupiahs). Thanks we revised the description including paragraph page 17, line 270.

In our study, COVID-19 cases were identified through positive RT-PCR test results, with tests conducted by Ministry of Health-accredited laboratories. We acknowledge the potential variability in access to PCR testing across different regions and times, as well as the possibility that inpatient testing may be more systematically documented than outpatient testing.

- The descriptors of cost were not presented. What types of costs were captured?

Our analysis encompassed direct healthcare-related costs in bundled payment, such as hospitalization, medication, and diagnostic tests, outpatient visits and follow-up care at health facilities based on the claim reimbursement. We amended the manuscript to include a clear type of cost were analysed.

- I don’t think using prior medical history to predict COVID-19 infection should be mentioned in the conclusions.

Thanks, deleted.

Attachment

Submitted filename: Response to Reviewer Feb24.docx

pone.0305835.s007.docx^{(22.1KB, docx)}

PLoS One. doi: 10.1371/journal.pone.0305835.r009

Decision Letter 4

Nemer Badwan

29 Feb 2024

PONE-D-22-31470R4Hospital services utilization and cost before and after COVID-19 hospital treatment: evidence from IndonesiaPLOS ONE

Dear Dr. Hafidz,

Please submit your revised manuscript by Apr 14 2024 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If applicable, we recommend that you deposit your laboratory protocols on protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions, see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind Regards,

Asst. Prof. Dr. Nemer Badwan

Ph.D in Economics and Finance

Assistant Professor of Economics and Finance

Academic Editor and Reviewer

PLOS ONE

[Note: HTML markup is below.]. [Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

Reviewer #5: All comments have been addressed

Reviewer #7: (No Response)

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

Reviewer #5: Yes

Reviewer #7: Partly

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #5: Yes

Reviewer #7: Yes

**********

4. Have the authors made all the data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with a rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information or deposited in a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g., participant privacy or use of data from a third party—those must be specified.

Reviewer #5: Yes

Reviewer #7: No

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

Reviewer #5: Yes

Reviewer #7: Yes

**********

6. Review Comments to the Author

Reviewer #5: Thanks for the answers to all my previous comments. I only have one additional comment.

Could you refer to all the S2 tables in the manuscript?

Reviewer #7: Review report on PONE-D-22-31470_R4: Hospital services utilization and cost before and after COVID-19 hospital treatment: evidence from Indonesia

The authors estimate health care utilization before and after PCR-confirmed SARS-CoV-2 in Indonesia in 2020 by contrasting the utilization of patients with COVID-19 in 2020 to the utilization of patients with COVID-19 in February 2021. They find elevated utilization around the time of infection, but no difference after four months.

The paper is well-written, and the statistical analyses appear well-executed. As noted by the authors, it is important to undertake studies using registry data in LMIC too, and this is a fine example thereof. Still, I believe the study suffers from a couple of fundamental concerns that need to be discussed more carefully and sincerely before the paper can be accepted for publication.

The construction of the treatment (covid-19) and comparison (not yet covid-19) is crucial for the interpretation of the results. The author states that “the comparison group comprised individuals treated for COVID-19 in hospitals in February 2021, under the presumption they had not contracted the virus in 2020” (line 115). This implies that what is estimated is not the impact (on subsequent utilization/costs) of having COVID-19 vs. not having COVID-19 in 2020, but the impact (on subsequent utilization/costs) of having COVID-19 in 2020 vs. having COVID-19 in 2021. This distinction has important implications for interpretation. For example, the authors’ DiD estimates how much higher the costs of treating a COVID-19 patient were in 2020 compared to 2021 (not the costs of treating a COVID-19 patient vs. not treating one). And, as another example, the authors do not shed any light on the presence of sequelae after SARS-CoV-2 infection, only on the question of whether possible sequelae are different in 2020 vs. 2021.

Clearly, these are two different questions (i. effects on utilization of COVID; ii. effects on utilization of COVID in 2020 vs. 2021), due, e.g., to the strain on health care services and other societal restrictions at the beginning of the pandemic. It seems obvious that the services’ ability to treat patients (both those with COVID-19 and other patients) was very different in 2020 than in 2021 (and definitely in 2024). It is crucial that the authors make this distinction between the two research questions clear to the readers (including in the abstract) and that they improve precision in how they describe their research question (and findings) throughout the paper (including considering making the title communicate this better).

In doing so, the authors may consider describing briefly other ways of constructing the comparison group and its implications for the research question, e.g., using contemporaneous patients with negative PCR tests (in 2020), like in reference (14), which would address the question of the impact of SARS-CoV-2 vs. no SARS-CoV-2 (instead of SARS-CoV-2 in 2020 vs. 2021). The authors should also be careful in pointing out the differences in research questions when comparing them with previous studies and when discussing policy implications (both in Discussion). This said, it appears to me that the question actually addressed by the authors' regressions (impact on utilization of COVID in 2020 vs. 2021) is of interest; it just needs to be stated and motivated clearly to the reader.

It seems that those dying are excluded from the sample, instead of the more common approach of censuring them from the month of death: “Similarly, those who died during COVID-19 hospitalization or in subsequent months were also excluded from the analysis.” (line 114). If medical treatment for COVID-19 was more effective in 2021 than in 2020 (e.g., due to less strain on the services or some immunity due to low-dose exposure or vaccination (too early for that in Indonesia?)), this introduces a potentially serious bias to the analysis: The comparison group receives more utilization because they survive longer (which would lead to an underestimation of how much higher costs were in 2020 than in 2021). This problem is not easy to handle, but it absolutely deserves a serious and sincere discussion. It would also be important to inform the reader of the survival in the two groups (i.e., by providing death rates by 3-6 months or even the impact on death using the same regressions). Was survival measured for the same length of time in the comparison group?

Minor things.

I do not understand why “individuals who were under monitoring and recorded as suspect or probable cases of COVID-19 in the previous months before laboratory confirmation were excluded” (line 113), since it makes the relevance of the results restricted to those who are actually being tested (which I presume could depend on the severity of the infection or socioeconomic status, or, as noted by the authors, that the hospital/patient have access to PCRs done by a Ministry of Health-accredited laboratory). On the other hand, I do not see that this has important methodological problems (except that PCR testing capacity was presumably smaller relative to the number of infections in 2020 than in February 2021) as long as the reader understands that the population under study is narrow. However, in note to Table 1, it is stated that “individuals without any hospital encounters before January 2020 (..) were excluded.” How is this incorporated for the comparison group in a way that does not introduce differences/bias across the two groups?

It would be very informative if Table 1 also included the outcome variables (i.e., the raw out- and inpatient rates and costs by, e.g., 3 months) and a reference to Supplement Table 2 for details (but this supplement table is not informative about the comparison group since we are only informed about post-infection months for the treated (not the comparison)—please fix).

Eq. 1 includes individual fixed effect (alpha i), but also a dummy for being in the treatment group (Covid i). I presume there is a typo here, as the individual fixed effect would absorb all other time-invariant individual characteristics (and, obviously, being in the treatment group does not change over time).

Given the vast population of Indonesia, some information on why there are so few people in the sample is warranted.

I don’t understand why the cost results are not also presented in a figure (like the nice figures for in- and outpatient results; please refer the reader to Supplement Table 1 in the note to these figures).

I was not able to reach the data application site using the link in the data sharing statement.

“before” is lacking in line 31 of the abstract.

Strictly speaking, the PCR test is for SARS-CoV-2 (i.e., the virus) and not for COVID-19 (the disease); the authors might consider their precision in the use of these two terms.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose "no," your identity will remain anonymous, but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #5: No

Reviewer #7: No

**********

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments.". [If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that supporting information files do not need this step.

PLoS One. 2024 Jul 5;19(7):e0305835. doi: 10.1371/journal.pone.0305835.r010

Author response to Decision Letter 4

22 Mar 2024

Reviewer 5

Thanks for the answers to all my previous comments. I only have one additional comment.

Could you refer to all the S2 tables in the manuscript?

We have now added the following sentences to refer to the Supplement S2 Tables in the end of our Results Section.

“In addition to the event-study estimates presented above, for transparency, we also reported the raw mean of monthly outpatient utilisation, inpatient utilisation, and hospital costs among COVID-19 groups and the comparison group (for all individuals) in Supplement S2 Tables B1, B2, and B3, respectively.”

Reviewer 7

Thank you for your comment. In our Methods, we have made clear that the Comparison group, those who were treated with COVID-19 in February 2021 were assumed to not contracted COVID-19 in 2020. In the Study Population subsection, we mentioned “Those who treated in hospitals with COVID-19 in February 2021 were classified as the control group, assuming they never contracted COVID-19 in 2020.”

We have added the following sentence to elaborate our assumption.

"This assumption is plausible given reinfection rate within one year follow up was only about 5% [17].”

In Study Population subsection, we revised our explanation to address your concern.

“We considered a period of August 2019 to November 2020 in our analysis. During this period, the comparison group, those who were treated with COVID-19 in February 2021, was assumed to not yet contracted the disease."

We believe that this is now clear that in our post-COVID follow-up, our comparison group was realistically assumed to have not contracted COVID-19, hence appropriate as control group and our main research question and subsequent interpretation of the results remain.

Thank you for your comment. First, we would like to reiterate that the comparison group (who were treated with COVID-19 in February 2021) had not yet contracted COVID during our pre- and post-COVID periods among those treated with COVID-19 in May to August 2020 (the COVID-19 group). This means that during these follow-up periods, the comparison group was assumed to use healthcare services that were not associated with them having contracted COVID-19.

Second, our decision to focus on survivors was precisely to avoid the issue of reduced healthcare use of COVID-19 patients due to their higher risk of mortality following their hospital treatment. We added the following explanation to our Study Population subsection.

“This decision was made to focus on COVID-19 survivors, hence avoiding potential issue of reduced healthcare use by the COVID-19 group due to their higher post-hospital treatment mortality.”

Minor things.

We agree with your concern that this is part of the data limitations. We have mentioned that the narrowness of this study population in our study limitation subsection.

Regarding the exclusion of individuals without any hospital encounters before January 2020, we made a mistake in the note from previous iteration of the analysis. The current results did not make this exclusion. We have revised our note to Table 1 as the following.

“Individuals who died of any causes during or after COVID-19 treatment were excluded.”

It would be difficult to provide the raw mean of our outcome variables in Table 1 for both COVID-19 and comparison group since the months that we used as pre- and post-COVID periods were different. In Figure 1, we have made it clear that, for example, in COVID-19 group who were hospitalised in May 2020, the post-COVID period that was analysed was June to November 2020.

The comparison of raw outcome means for COVID-19 groups and the comparison group were provided in S2 Tables. To make the Tables more informative, we have now revised the Table notes as the following.

“The shaded cells are the post-COVID period for each COVID-19 group. The underlined figures are the reference period for each COVID-19 group. A simple analysis is to compare the outcomes in the same months of the COVID-19 groups with the comparison group, adjusted with the difference in outcome between the two groups at the reference month (the underlined figure). For example, the post-COVID period for May 2020 COVID-19 group is June to November 2020 and the reference month is December 2019. Robust standard errors in parentheses.”

Thank you for the correction. We have now deleted α_i from the equation since it is not estimated in our regression.

Given the vast population of Indonesia, some information on why there are so few people in the sample is warranted.

We have now revised the first limitation of the paper as the following.

“First, the study population was limited to NHI members who had at least one regular hospital encounter before being treated with laboratory-confirmed COVID-19. This made our sample smaller than the cumulative number of confirmed cases in Indonesia which were about 1.3 million cases as of February 2021.”

We did not present the cost results as Figures because the scale of the estimated effect of COVID-19 across period greatly differed. For example, in Table 2, the point estimate for May 2020 COVID-19 group at 1 month after COVID-19 hospitalisation was 2,453 (in thousand IDR) while the point estimate at the hospitalisation month was 168,649 (in thousand IDR). Because of this concern, presenting the results as Figures would not be as informative as detailed results provided in Tables 2-4.

I was not able to reach the data application site using the link in the data sharing statement.

We are aware about the difficulty of accessing BPJS Kesehatan website from overseas. This issue is, however, beyond our capacity as researchers.

“before” is lacking in line 31 of the abstract.

Thank you for pointing this out. We have now revised this.

Strictly speaking, the PCR test is for SARS-CoV-2 (i.e., the virus) and not for COVID-19 (the disease); the authors might consider their precision in the use of these two terms.

Thank you for the suggestion. We have clearly stated in our Study Population section that we define COVID-19 in our paper as PCR-confirmed COVID-19 cases. We believe that this narrower definition is appropriate to avoid too many clinical practice variations in COVID-19 diagnosis, especially in the early years of the pandemic.

Attachment

Submitted filename: Response to reviewer.docx

pone.0305835.s008.docx^{(20.6KB, docx)}

PLoS One. doi: 10.1371/journal.pone.0305835.r011

Decision Letter 5

Nemer Badwan

17 Apr 2024

PONE-D-22-31470R5Hospital services utilisation and cost before and after COVID-19 hospital treatment: evidence from IndonesiaPLOS ONE

Dear Dr. Hafidz,

==============================

Dear author(s),

The seventh reviewer asked to amend all the main comments he requested in the fourth review round, and you did not amend what was required.

Either you are not convinced by those comments and are against them, or you did not fully understand what the reviewer wanted from you in order to amend it as necessary.

Therefore, I ask you to review the comments of the seventh reviewer, amend and include all the comments and concerns raised by the reviewer, and submit the manuscript for the sixth time for review.

Please carefully review all comments and do not ignore any of the comments already included in the previous peer review report.

Please see below the reviewer's 7 comments:

I'm not impressed by the authors' response to my comments and suggestions. They either do not want to understand my two main comments—and if so, they should say so or state that they disagree with me (which could be fine, of course)—or their understanding of the method they are applying (DID) is very shallow.

In particular, in responding to my first main comment on implications for interpretations of using 2021 cases as a comparison group, they keep repeating that they assume that the comparison group had not had COVID before. While I agree that this is an important assumption, my point was another one: The DID they run looks at how changes from before to after COVID differ for those having it in 2020 vs. 2021; see my prior report to the authors for elaboration.

I'm also not impressed with the shallow response to my second main comment.

Overall, I maintain my comment in my previous report that the interpretations they make do not sufficiently reflect the research question actually addressed by the method they in fact applied. As noted in my previous response to you, I do not think there is a need for new analyses, only that the interpretation and framing of the paper should align better with what the applied method (DID) is in fact addressing.

On the other hand, it seems to me that the authors have had to deal with 7 (?!) reviewers and a number of revisions, so I can understand if they are getting fed up with revising this paper. Thus, I see that making a decision on this paper can be a hard call for the editor.

==============================

Please submit your revised manuscript by Jun 01 2024 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

We look forward to receiving your revised manuscript.

Kind Regards,

Asst. Prof. Dr. Nemer Badwan

Ph.D in Economics and Finance

Academic Editor and Reviewer

PLOS ONE

Journal Requirements:

Additional Editor Comments (if provided):

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

Reviewer #5: All comments have been addressed

Reviewer #7: (No Response)

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

Reviewer #5: Yes

Reviewer #7: Partly

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #5: Yes

Reviewer #7: I Don't Know

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

Reviewer #5: Yes

Reviewer #7: No

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

Reviewer #5: Yes

Reviewer #7: Yes

**********

6. Review Comments to the Author

Reviewer #5: (No Response)

Reviewer #7: (No Response)

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #5: No

Reviewer #7: No

**********

PLoS One. 2024 Jul 5;19(7):e0305835. doi: 10.1371/journal.pone.0305835.r012

Author response to Decision Letter 5

16 May 2024

Response to Reviewer #7

Thanks for your reply to our responses in the previous review report. We would like to elaborate more on them as the reviewer appears to see our previous replies to be insufficient.

First, what we understand is that your first main concern is related to the application and interpretation of our DiD event-study approach. You concluded that we estimated the “effects on utilisation of COVID in 2020 vs 2021”. We would like to clarify that this is not what we meant.

As illustrated in Figure 1, we conducted four separate DiD event-study estimations, for cohort of individuals who were treated with COVID in May, June, July, and August 2020. We would like to explain it as clearly as possible here using the example of the event-study which examining the impact of COVID-19 for those hospitalised with SARS-CoV-2 in May 2020 (the top left panel of Figure 3).

As stated in our statistical analysis, “we used the fifth month leading to the hospitalisation as the reference period.”

New text: “For example, May 2020 group, this reference period was December 2019 and labelled as t-5. The t0 was May 2020, the t+1 to t+6 was June to November 2020, the t-4 to t-1 was January 2020 to April 2020. For each subsequent month, the reference and comparative periods shifted accordingly”.

Overall, it appears that we have disagreements around the implementation and interpretation of the DiD approach in our analysis. First, the interpretation of estimates before and after COVID-19 is from the perspective of those who were treated with COVID-19 between May and August 2020, in comparison to those who "will" eventually get COVID-19 in February 2021. In this case, we believe that our explanation is clear enough to highlight that the comparison group has not contracted COVID-19 in the analytical period.

New Text in Study Population Section: " The comparison group comprised individuals treated for COVID-19 in hospitals in February 2021 (not hospitalised in 2020), under the assumption they had not contracted the virus in 2020. We argue this is a suitable comparator for our analysis as these groups are similar (i.e. attitude towards risk, preference, etc.). As we observed their health care utilisation in the entirety of our estimation sample from August 2019 to November 2020, the event study approach allows us to compare the outcomes between treatment and comparison groups over time. "

As for your second main comment, we apologise if we did not respond to it as detailed as you expected, especially the last part of your second comment ("It would also be important to inform the reader of the survival in the two groups (i.e., by providing death rates by 3-6 months or even the impact on death using the same regressions). Was survival measured for the same length of time in the comparison group?"). We would like to answer it as the following:

We agree that survival is an important factor that determines healthcare costs and utilisations. We have answered this concern by stating that our focus is on the survivors. Again, from the perspective of the COVID-19 group, within the analytical periods (August 2019 to November 2020), all individuals in the comparison group had not contracted COVID-19 yet. Therefore, no survival rate associated with COVID-19 can be inferred from the comparison group. To provide some context in terms of the recorded mortality of the COVID-19 group, we have added the following sentence to our Study Population subsection.

“This decision was made to focus on COVID-19 survivors, hence avoiding the potential issue of reduced healthcare use by the COVID-19 group due to their higher post-hospital treatment mortality. For context, the recorded 6-month mortality rate of those treated with COVID-19 between May and August 2020 was 12.1%”

PLoS One. doi: 10.1371/journal.pone.0305835.r013

Decision Letter 6

Nemer Badwan

6 Jun 2024

Hospital services utilisation and cost before and after COVID-19 hospital treatment: evidence from Indonesia

PONE-D-22-31470R6

Dear Dr. Firdaus Hafidz,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice will be generated when your article is formally accepted. Please note, if your institution has a publishing partnership with PLOS and your article meets the relevant criteria, all or part of your publication costs will be covered. Please make sure your user information is up-to-date by logging into Editorial Manager at Editorial Manager® and clicking the ‘Update My Information' link at the top of the page. If you have any questions relating to publication charges, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind Regards,

Asst. Prof. Dr. Nemer Badwan

Ph.D in Economics and Finance

Assistant Professor of Economics and Finance

Academic Editor and Reviewer

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

Reviewer #5: All comments have been addressed

Reviewer #8: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

Reviewer #5: Yes

Reviewer #8: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #5: Yes

Reviewer #8: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

Reviewer #5: Yes

Reviewer #8: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

Reviewer #5: Yes

Reviewer #8: Yes

**********

6. Review Comments to the Author

Reviewer #5: I have reviewed this manuscript before, and it has improved during the revisions. I have no further comments.

Reviewer #8: Dear authors,

thank you for this opportunity to read "Hospital services utilisation and cost before and after COVID-19 hospital treatment: evidence from Indonesia". It was very clear and well written.

My only concern, and it is minor and should not delay publication, is that in future articles, please be clear about the average outpatient visit costs or per capita expenditure, by island, age, and gender, should be if people were in good health. Reviewing your supplemental tables assisted in my understanding of your paper. Reporting that information will help future readers understand the magnitude and burden of increased utilization on the Indonesian health system.

Overall, a good paper. I look forward to seeing it in print.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #5: No

Reviewer #8: No

**********

PLoS One. doi: 10.1371/journal.pone.0305835.r014

Acceptance letter

Nemer Badwan

25 Jun 2024

PONE-D-22-31470R6

PLOS ONE

Dear Dr. Hafidz,

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now being handed over to our production team.

At this stage, our production department will prepare your paper for publication. This includes ensuring the following:

* All references, tables, and figures are properly cited

* All relevant supporting information is included in the manuscript submission,

* There are no issues that prevent the paper from being properly typeset

If revisions are needed, the production department will contact you directly to resolve them. If no revisions are needed, you will receive an email when the publication date has been set. At this time, we do not offer pre-publication proofs to authors during production of the accepted work. Please keep in mind that we are working through a large volume of accepted articles, so please give us a few weeks to review your paper and let you know the next and final steps.

Lastly, if your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

If we can help with anything else, please email us at customercare@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Asst. Prof. Dr. Nemer Badwan

Academic Editor

PLOS ONE

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

S1 Table. Summary of dataset used in the manuscript to create Figs 3 and 4.

(PDF)

pone.0305835.s001.pdf^{(228.8KB, pdf)}

S2 Table. Crude hospital utilisation rate and cost over time.

(PDF)

pone.0305835.s002.pdf^{(122.1KB, pdf)}

Attachment

Submitted filename: PONE-D-22-31470_reviewer.pdf

pone.0305835.s003.pdf^{(1.3MB, pdf)}

Attachment

Submitted filename: Responses to Reviewers.docx

pone.0305835.s004.docx^{(46.4KB, docx)}

Attachment

Submitted filename: Response to Reviewer Sept.docx

pone.0305835.s005.docx^{(24.2KB, docx)}

Attachment

Submitted filename: Response to Reviewer Sept.docx

pone.0305835.s006.docx^{(24.2KB, docx)}

Attachment

Submitted filename: Response to Reviewer Feb24.docx

pone.0305835.s007.docx^{(22.1KB, docx)}

Attachment

Submitted filename: Response to reviewer.docx

pone.0305835.s008.docx^{(20.6KB, docx)}

Data Availability Statement

[pone.0305835.ref001] 1.Cascella M, Rajnik M, Aleem A, Dulebohn SC, Di Napoli R. Features, Evaluation, and Treatment of Coronavirus (COVID-19). StatPearls. Treasure Island (FL); 2023. [PubMed] [Google Scholar]

[pone.0305835.ref002] 2.Li J, Huang DQ, Zou B, Yang H, Hui WZ, Rui F, et al. Epidemiology of COVID-19: A systematic review and meta-analysis of clinical characteristics, risk factors, and outcomes. J Med Virol. 2021;93(3):1449–58. Epub 2020/08/14. doi: 10.1002/jmv.26424 . [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0305835.ref003] 3.Stokes EK, Zambrano LD, Anderson KN, Marder EP, Raz KM, El Burai Felix S, et al. Coronavirus Disease 2019 Case Surveillance—United States, January 22-May 30, 2020. MMWR Morb Mortal Wkly Rep. 2020;69(24):759–65. Epub 2020/06/20. doi: 10.15585/mmwr.mm6924e2 . [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0305835.ref004] 4.Surendra H, Elyazar IR, Djaafara BA, Ekawati LL, Saraswati K, Adrian V, et al. Clinical characteristics and mortality associated with COVID-19 in Jakarta, Indonesia: A hospital-based retrospective cohort study. Lancet Reg Health West Pac. 2021;9:100108. Epub 2021/03/09. doi: 10.1016/j.lanwpc.2021.100108 . [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0305835.ref005] 5.Surendra H, Paramita D, Arista NN, Putri AI, Siregar AA, Puspaningrum E, et al. Geographical variations and district-level factors associated with COVID-19 mortality in Indonesia: a nationwide ecological study. BMC Public Health. 2023;23(1):103. Epub 2023/01/15. doi: 10.1186/s12889-023-15015-0 . [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0305835.ref006] 6.Taquet M, Dercon Q, Luciano S, Geddes JR, Husain M, Harrison PJ. Incidence, co-occurrence, and evolution of long-COVID features: A 6-month retrospective cohort study of 273,618 survivors of COVID-19. PLoS Med. 2021;18(9):e1003773. Epub 2021/09/29. doi: 10.1371/journal.pmed.1003773 . [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0305835.ref007] 7.Chen C, Haupert SR, Zimmermann L, Shi X, Fritsche LG, Mukherjee B. Global Prevalence of Post COVID-19 Condition or Long COVID: A Meta-Analysis and Systematic Review. J Infect Dis. 2022. Epub 2022/04/17. doi: 10.1093/infdis/jiac136 . [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0305835.ref008] 8.Ramzi ZS. Hospital readmissions and post-discharge all-cause mortality in COVID-19 recovered patients; A systematic review and meta-analysis. Am J Emerg Med. 2022;51:267–79. Epub 2021/11/16. doi: 10.1016/j.ajem.2021.10.059 . [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0305835.ref009] 9.Humphreys H, Kilby L, Kudiersky N, Copeland R. Long COVID and the role of physical activity: a qualitative study. BMJ Open. 2021;11(3):e047632. Epub 2021/03/12. doi: 10.1136/bmjopen-2020-047632 . [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0305835.ref010] 10.Nandasena H, Pathirathna ML, Atapattu A, Prasanga PTS. Quality of life of COVID 19 patients after discharge: Systematic review. PLoS One. 2022;17(2):e0263941. Epub 2022/02/17. doi: 10.1371/journal.pone.0263941 . [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0305835.ref011] 11.Nopp S, Moik F, Klok FA, Gattinger D, Petrovic M, Vonbank K, et al. Outpatient Pulmonary Rehabilitation in Patients with Long COVID Improves Exercise Capacity, Functional Status, Dyspnea, Fatigue, and Quality of Life. Respiration. 2022;101(6):593–601. Epub 2022/02/25. doi: 10.1159/000522118 . [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0305835.ref012] 12.Menges D, Ballouz T, Anagnostopoulos A, Aschmann HE, Domenghino A, Fehr JS, et al. Burden of post-COVID-19 syndrome and implications for healthcare service planning: A population-based cohort study. PLoS One. 2021;16(7):e0254523. Epub 2021/07/13. doi: 10.1371/journal.pone.0254523 . [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0305835.ref013] 13.Ayoubkhani D, Khunti K, Nafilyan V, Maddox T, Humberstone B, Diamond I, et al. Post-covid syndrome in individuals admitted to hospital with covid-19: retrospective cohort study. BMJ. 2021;372:n693. Epub 2021/04/02. doi: 10.1136/bmj.n693 . [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0305835.ref014] 14.Magnusson K, Skyrud KD, Suren P, Greve-Isdahl M, Stordal K, Kristoffersen DT, et al. Healthcare use in 700 000 children and adolescents for six months after covid-19: before and after register based cohort study. BMJ. 2022;376:e066809. Epub 2022/01/19. doi: 10.1136/bmj-2021-066809 . [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0305835.ref015] 15.Skyrud KD, Hernaes KH, Telle KE, Magnusson K. Impacts of mild COVID-19 on elevated use of primary and specialist health care services: A nationwide register study from Norway. PLoS One. 2021;16(10):e0257926. Epub 2021/10/09. doi: 10.1371/journal.pone.0257926 . [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0305835.ref016] 16.Agustina R, Dartanto T, Sitompul R, Susiloretni KA, Suparmi, Achadi EL, et al. Universal health coverage in Indonesia: concept, progress, and challenges. The Lancet. 2019;393(10166):75–102. doi: 10.1016/s0140-6736(18)31647-7 [DOI] [PubMed] [Google Scholar]

[pone.0305835.ref017] 17.Medic S, Anastassopoulou C, Lozanov-Crvenkovic Z, Vukovic V, Dragnic N, Petrovic V, et al. Risk and severity of SARS-CoV-2 reinfections during 2020–2022 in Vojvodina, Serbia: A population-level observational study. Lancet Reg Health Eur. 2022;20:100453. Epub 20220701. doi: 10.1016/j.lanepe.2022.100453 . [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0305835.ref018] 18.Hendarwan H, Syachroni S, Aryastami NK, Su’udi A, Susilawati MD, Despitasari M, et al. Assessing the COVID-19 diagnostic laboratory capacity in Indonesia in the early phase of the pandemic. WHO South East Asia J Public Health. 2020;9(2):134–40. Epub 2020/09/27. doi: 10.4103/2224-3151.294307 . [DOI] [PubMed] [Google Scholar]

[pone.0305835.ref019] 19.Caniglia EC, Murray EJ. Difference-in-Difference in the Time of Cholera: a Gentle Introduction for Epidemiologists. Curr Epidemiol Rep. 2020;7(4):203–11. Epub 2021/04/02. doi: 10.1007/s40471-020-00245-2 . [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0305835.ref020] 20.Wing C, Simon K, Bello-Gomez RA. Designing Difference in Difference Studies: Best Practices for Public Health Policy Research. Annu Rev Public Health. 2018;39:453–69. Epub 2018/01/13. doi: 10.1146/annurev-publhealth-040617-013507 . [DOI] [PubMed] [Google Scholar]

[pone.0305835.ref021] 21.Rothbard S, Etheridge JC, Murray EJ. A Tutorial on Applying the Difference-in-Differences Method to Health Data. Current Epidemiology Reports. 2023. doi: 10.1007/s40471-023-00327-x [DOI] [Google Scholar]

[pone.0305835.ref022] 22.Driver JA, Djousse L, Logroscino G, Gaziano JM, Kurth T. Incidence of cardiovascular disease and cancer in advanced age: prospective cohort study. BMJ. 2008;337:a2467. Epub 2008/12/11. doi: 10.1136/bmj.a2467 . [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0305835.ref023] 23.Lavery AM, Preston LE, Ko JY, Chevinsky JR, DeSisto CL, Pennington AF, et al. Characteristics of hospitalized COVID-19 patients discharged and experiencing same-hospital readmission—United States, March–August 2020. Morbidity and Mortality Weekly Report. 2020;69(45):1695. doi: 10.15585/mmwr.mm6945e2 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0305835.ref024] 24.Donnelly JP, Wang XQ, Iwashyna TJ, Prescott HC. Readmission and Death After Initial Hospital Discharge Among Patients With COVID-19 in a Large Multihospital System. JAMA. 2021;325(3):304–6. Epub 2020/12/15. doi: 10.1001/jama.2020.21465 . [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0305835.ref025] 25.Iannaccone S, Alemanno F, Houdayer E, Brugliera L, Castellazzi P, Cianflone D, et al. COVID-19 rehabilitation units are twice as expensive as regular rehabilitation units. J Rehabil Med. 2020;52(6):jrm00073. Epub 2020/06/10. doi: 10.2340/16501977-2704 . [DOI] [PubMed] [Google Scholar]

[pone.0305835.ref026] 26.Lund LC, Hallas J, Nielsen H, Koch A, Mogensen SH, Brun NC, et al. Post-acute effects of SARS-CoV-2 infection in individuals not requiring hospital admission: a Danish population-based cohort study. The Lancet Infectious Diseases. 2021;21(10):1373–82. doi: 10.1016/S1473-3099(21)00211-5 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0305835.ref027] 27.Cohen K, Ren S, Heath K, Dasmarinas MC, Jubilo KG, Guo Y, et al. Risk of persistent and new clinical sequelae among adults aged 65 years and older during the post-acute phase of SARS-CoV-2 infection: retrospective cohort study. BMJ. 2022;376:e068414. Epub 2022/02/11. doi: 10.1136/bmj-2021-068414 . [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0305835.ref028] 28.Gao YD, Ding M, Dong X, Zhang JJ, Kursat Azkur A, Azkur D, et al. Risk factors for severe and critically ill COVID-19 patients: A review. Allergy. 2021;76(2):428–55. Epub 2020/11/14. doi: 10.1111/all.14657 . [DOI] [PubMed] [Google Scholar]

[pone.0305835.ref029] 29.van Empel G, Mulyanto J, Wiratama BS. Undertesting of COVID-19 in Indonesia: what has gone wrong? J Glob Health. 2020;10(2):020306. Epub 2020/10/29. doi: 10.7189/jogh.10.020306 . [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0305835.ref030] 30.Ariawan I, Jusril H. COVID-19 in Indonesia: where are we? Acta Medica Indonesiana. 2020;52(3):193. [PubMed] [Google Scholar]

[pone.0305835.ref031] 31.Djalante R, Lassa J, Setiamarga D, Sudjatma A, Indrawan M, Haryanto B, et al. Review and analysis of current responses to COVID-19 in Indonesia: Period of January to March 2020. Prog Disaster Sci. 2020;6:100091. Epub 2020/04/01. doi: 10.1016/j.pdisas.2020.100091 . [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0305835.ref032] 32.Bellelli G, Rebora P, Valsecchi MG, Bonfanti P, Citerio G, members C-MT. Frailty index predicts poor outcome in COVID-19 patients. Intensive Care Med. 2020;46(8):1634–6. Epub 20200525. doi: 10.1007/s00134-020-06087-2 . [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0305835.ref033] 33.Nugraha RR, Pratiwi MA, Al-Faizin RE, Permana AB, Setiawan E, Farianty Y, et al. Predicting the cost of COVID-19 treatment and its drivers in Indonesia: analysis of claims data of COVID-19 in 2020–2021. Health Econ Rev. 2022;12(1):45. Epub 2022/09/01. doi: 10.1186/s13561-022-00392-w . [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0305835.ref034] 34.Ministry of Health. PETUNJUK TEKNIS KLAIM PENGGANTIAN BIAYA PERAWATAN PASIEN PENYAKIT INFEKSI EMERGING TERTENTU BAGI RUMAH SAKIT YANG MENYELENGGARAKAN PELAYANAN CORONAVIRUS DISEASE 2019. Jakarta2020.

[pone.0305835.ref035] 35.Dagens A, Sigfrid L, Cai E, Lipworth S, Cheng V, Harris E, et al. Scope, quality, and inclusivity of clinical guidelines produced early in the covid-19 pandemic: rapid review. BMJ. 2020;369:m1936. Epub 2020/05/28. doi: 10.1136/bmj.m1936 . [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0305835.ref036] 36.Whittaker HR, Gulea C, Koteci A, Kallis C, Morgan AD, Iwundu C, et al. GP consultation rates for sequelae after acute covid-19 in patients managed in the community or hospital in the UK: population based study. BMJ. 2021;375:e065834. Epub 2021/12/31. doi: 10.1136/bmj-2021-065834 . [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0305835.ref037] 37.Arifin B, Anas T. Lessons learned from COVID-19 vaccination in Indonesia: experiences, challenges, and opportunities. Hum Vaccin Immunother. 2021;17(11):3898–906. Epub 2021/10/07. doi: 10.1080/21645515.2021.1975450 . [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Hospital services utilisation and cost before and after COVID-19 hospital treatment: Evidence from Indonesia

Muhammad Fikru Rizal

Firdaus Hafidz

Gilbert Renardi Kusila

Wan Aisyiah

Dedy Revelino

Erzan Dhanalvin

Ayunda Oktavia

Ilyasa

Citra Jaya

Benjamin Saut

Mahlil Ruby

Roles

Abstract

Objective

Methods

Results

Conclusion

Introduction

Methods

Data source

Ethical clearance

Study population

Study period

Fig 1. Observation period for COVID-19 and comparison group.

Outcome variables

Statistical analysis

Results

Study participants

Fig 2. Study participants flow.

Table 1. Baseline characteristics of COVID-19 cases treated in hospitals in Indonesia from May to August 2020 compared with the comparison group.

Changes in utilisation of hospital services

Fig 3. Difference-in-difference estimates of hospital outpatient visit rate of COVID-19 group compared to comparison group.

Fig 4. Difference-in-difference estimates of hospital inpatient rate of COVID-19 group compared to comparison group.

Changes in total costs of hospital services

Table 2. Difference-in-difference estimates of total costs of hospital services of COVID-19 group compared to comparison group.

Table 4. Difference-in-difference estimates of total costs of hospital services of COVID-19 group compared to comparison group by gender.

Table 3. Difference-in-difference estimates of total costs of hospital services of COVID-19 group compared to comparison group by age.

Discussion

Main findings

Topics for future study

Strengths and limitations of this study

Policy implication

Conclusion

Supporting information

Acknowledgments

Data Availability

Funding Statement

References

Decision Letter 0

Nemer Badwan

Roles

Author response to Decision Letter 0

Decision Letter 1

Nemer Badwan

Roles

Author response to Decision Letter 1

Decision Letter 2

Nemer Badwan

Roles

Author response to Decision Letter 2

Decision Letter 3

Nemer Badwan

Roles

Author response to Decision Letter 3

Decision Letter 4

Nemer Badwan

Roles

Author response to Decision Letter 4

Decision Letter 5

Nemer Badwan

Roles

Author response to Decision Letter 5

Decision Letter 6

Nemer Badwan

Roles

Acceptance letter

Nemer Badwan

Roles