Costs and Late Complications of Pediatric Appendicitis

Abstract

Background

Immediate complications of appendicitis are common, but the prevalence of long-term complications is uncertain.

Methods

We studied all publicly-insured children in the US with appendicitis in 2018–2019 using administrative claims. The main outcome was late hospital care defined as hospitalization or abdominal procedure within 180 days of an appendicitis discharge, excluding interval appendectomies. Time to late hospital care was evaluated using Cox regression. We evaluated healthcare expenditures arising from appendicitis episodes.

Results

Among 95,942 children with appendicitis, 5,727 (6.0%) had late hospital care, with 5,062 requiring rehospitalization and 2,012 (2.1%) surgery. The median time to late hospital care was 10 days (IQR 4–33). Age under 5 years (compared with >14 years, hazard ratio [HR] 1.88, 95% confidence interval [CI] 1.70–2.08), complex chronic conditions (HR 2.35, 95% CI 2.13–2.59), and complicated appendicitis (HR 2.81, 95% CI 2.67, 2.96) were each associated with time to late hospital care. Expenditures over 180 days were a median $6,553 and $19,589 respectively in those requiring no late hospital care versus those requiring it (p<0.001).

Conclusions

Late hospital care is uncommon in pediatric appendicitis but is costly. Prevention efforts should be targeted to the youngest, most complex children, and those with complicated appendicitis at presentation.

Introduction

Appendicitis is the most common serious surgical emergency in children.¹ Immediate complications are common, occurring about 25% of the time, and include perforated appendicitis, sepsis, and rarely a need for bowel resection.² However, it is less clear how frequently later complications occur after the index episode has concluded. A marker for late complications is the need for subsequent hospital care, including rehospitalization or surgery other than interval appendectomy.

Late hospital care is an important outcome because it represents ongoing and protracted illness which impacts patients’ lives. For example, the need for later surgery after an appendicitis diagnosis would imply that symptoms were ongoing or that the illness did not resolve. In addition, late hospital care impacts overall cost of care, although the magnitude of such cost increases is unclear. Knowledge of the frequency and predictors of late hospital care could inform prevention efforts. In addition, it is unclear whether research assessing outcomes of pediatric appendicitis needs to account for care beyond the initial diagnosis or hospitalization.

To address these issues, using a national population of publicly insured children, we sought to evaluate rates, predictors, and costs of pediatric appendicitis according to whether complications occurred after the index encounter.

Methods

We performed a retrospective cohort study of all United States children enrolled in Medicaid or the Children’s Health Insurance Program in 2018 or 2019 with a first-time diagnosis of appendicitis in an emergency department (ED) or inpatient unit. Eligible patients were identified in the Transformed Medicaid Statistical Information System Analytic Files (TAF).³ The TAF includes all claims for individuals enrolled in Medicaid and the Children’s Health Insurance Program. Appendicitis encounters were identified based on an International Classification of Diseases 10th Edition (ICD-10) Clinical Modification diagnosis code for appendicitis (K35-K37) in an ED visit. ED visits were identified based on a strategy used in analogous Medicare claims (Supplemental Methods).⁴ We excluded children from Illinois where institutional claims are not reliably coded, those who died during the index encounter, and patients whose Medicaid enrollment ended before the index appendicitis encounter concluded.

The primary outcome was time to the first episode of late hospital care after an appendicitis encounter. All late hospital care occurred after discharge from the index encounter, defined as the index hospitalization and/or surgery. Late hospital care included all-cause rehospitalization or subsequent surgery excluding interval appendectomy occurring within 180 days of the appendicitis encounter. Interval appendectomies were defined as a first appendectomy occurring after the index appendicitis encounter.⁵ For example, an appendectomy and its associated hospitalization occurring 2 months after an appendicitis diagnosis would be considered an interval appendectomy if no appendectomy occurred at the time of appendicitis diagnosis. Surgeries were identified using ICD-10 codes, Current Procedural Terminology codes, or an operating room charge.⁶ Time to late hospital care was measured from the end of the index appendicitis encounter to the start of the encounter in which late care occurred. For patients without late hospital care, the time interval was censored at the earlier of the patient’s public insurance eligibility, the end of the study, or 180 days (as we felt care outside of this window would be highly unlikely to be related to the index appendicitis episode).

Secondary outcomes included overall 30- and 180-day late complication rates. We also measured 30- and 180-day total true costs of an appendicitis episode which were the sum of all actual healthcare expenditures by the insurer or patient. Costs included all care, including outpatient, ED, and inpatient expenditures, but did not include pharmacy costs which were not available in the database. The 30- or 180-day interval started at the end of the index appendicitis encounter. For cost analyses, only patients with a fee-for-service insurance plan or plans that reported fee-for-service equivalents were analyzed, since expenditures are often unreported for patients on capitated plans.

Covariates included patient demographics including age, sex, primary language, complex chronic condition⁷, and socioeconomic status measured using the child opportunity index at the ZIP code level.⁸ We determined the presence of complicated appendicitis using diagnosis codes⁹.

Analysis

We reported demographics and follow-up time distributions using descriptive statistics. We next determined overall, 30-, and 180-day late hospital care rates and types. For late hospitalizations, we reported the primary diagnoses categorized using Clinical Classifications Software.¹⁰ Associations of covariates with time to late hospital care were assessed using bivariable Cox regressions. Cox models were used because of the varying follow-up times by patient. Kaplan-Meier plots were generated for covariates significant at p < 0.05 in the bivariable models. Adjusted risks of earlier time to late hospital care were evaluated using a single Cox model including all covariates. Each covariate was included because we hypothesized that each could plausibly contribute to the likelihood and timing of late hospital care.

We next analyzed costs, using only patients with continuous enrollment and available data for 30 days (for 30-day costs) or 180 days (for 180-day costs). We reported total 30- and 180-day costs according to whether late hospital care occurred. Costs were also determined at the index versus later encounters. Costs were attributed to ED/inpatient or outpatient care based on each day’s expenditures (i.e. an expense during an inpatient encounter was categorized as an inpatient expense even if it was on a separate claim, which frequently occurs in professional bills). Finally, to evaluate the extent to which late complications account for late costs, we performed hierarchical linear regression with the dependent variable of 30- or 180-day costs and a random intercept for hospital nested within state and fixed effects including all covariates and late hospital care.

This study was certified as exempt from review by the Institutional Review Board.

Results

There were 99,851 eligible patients with appendicitis. We excluded 3,003 (3.0%) from Illinois, 13 (<0.1%) who died during the index encounter, and 893 (0.9%) for enrollment ending before the index encounter concluded. We therefore analyzed 95,942 (96.1%) patients. Follow-up time was a complete 180 days for 67,403 (70.3%) of patients and at least 30 days for 90,453 (94.3%). A plurality of patients was 10–14 years old, male, located in low-opportunity areas, had uncomplicated appendicitis, and did not have a CCC (Table 1). Interval appendectomy was performed in 1,531 (1.6%) patients.

Table 1:

Demographics features of the cohort.

	N=95,953 n (%)
Age, years
<5	6122 (6.4)
5–9	30331 (31.6)
10–14	40824 (42.6)
>14	18665 (19.5)
Female	38038 (39.6)
Child opportunity index quartile
Lowest	35703 (37.6)
2nd	29996 (31.6)
3rd	19629 (20.6)
Highest	9742 (10.2)
Complex chronic condition	3390 (3.5)
Complicated appendicitis	26349 (27.5)
Index appendectomy	68909 (71.8)
Interval appendectomy*	1531 (1.6)

^*Interval appendectomy was defined as the first use of a procedure code for appendectomy after the index encounter, as long as the index encounter did not include an appendectomy. Hospitalizations associated with interval appendectomy were not considered late complications. Patients with interval appendectomy could still have the primary outcomes of late hospital care if they had additional surgeries or hospitalizations other than the interval appendectomy.

Late hospital care occurred in 4195 (4.4%) patients within 30 days and 5,727 (6.0%) patients within 180 days. Late hospitalizations occurred in 5,062 (5.3%) patients and 2,891 late surgeries were performed in 2,012 (2.1%) patients. The index encounter length of stay was a median 1 (IQR 1, 2) day among those without late complications and 2 (IQR 0, 5) among those with late complications (p<0.001). The most common primary diagnoses among late hospitalizations were appendicitis and other appendiceal conditions (15.2%), complications of medical or surgical care (14.8%), abdominal pain (9.8%), peritonitis and intra-abdominal abscess (8.9%), and intestinal obstruction and ileus (5.0%) (Supplemental Table 1). Late surgeries were categorized as abdominal drainage (31.8% of all late surgeries), repeat appendectomy (i.e. any appendectomy performed after an initial or interval appendectomy, 13.4% of all late surgeries, or 0.3% of all appendicitis patients), bowel/omentum release (8.4%), bowel resection (5.6%), peritoneal inspection (2.6%), other (2.2%), and unknown (based on the presence of an operating room charge without a specific procedure code: 36.0%). Among children with complicated appendicitis, 2,878 (10.9%) had late hospital care.

The median time to late hospital care was 10 days (IQR 4, 33) from discharge after the index encounter. In bivariable models, age < 5 years compared with >14 (hazard ratio [HR] 1.88, 95% confidence interval [CI] 1.70, 2.08), CCC (HR 2.35, 95% CI 2.13, 2.59), and complicated appendicitis (HR 2.81, 95% CI 2.67, 2.96) were each associated with a shorter time to late hospital care (Table 2). Time to late hospital care is shown by significant covariate from bivariable models in Figure 1. In the adjusted model, children living in the highest-opportunity areas had a shorter time to late hospital care.

Table 2:

Rates of late hospital care by risk factor. Unadjusted (one variable in model) and adjusted (all variables in model) hazard ratios for time to late hospital care are shown.

Variable	Late hospital care n/N (%)	Unadjusted hazard ratio (95% CI)	Adjusted hazard ratio (95% CI)
Age
<5	614/6122 (10.0)	1.88 (1.70, 2.08)	1.34 (1.21, 1.49)
10–14	1744/30331 (5.7)	1.05 (0.97, 1.13)	0.94 (0.87, 1.02)
5–9	2355/40824 (5.8)	1.05 (0.98, 1.13)	0.98 (0.91, 1.06)
>14	1014/18665 (5.4)	Referent	Referent
Complex chronic condition
No	5294/92552 (5.7)	Referent	Referent
Yes	433/3390 (12.8)	2.35 (2.13, 2.59)	2.49 (2.26, 2.75)
Child opportunity index quartile
Lowest	2124/35703 (5.9)	Referent	Referent
2nd	1726/29996 (5.8)	0.97 (0.91, 1.03)	1.00 (0.94, 1.07)
3rd	1208/19629 (6.2)	1.04 (0.97, 1.11)	1.08 (1.01, 1.16)
Highest	621/9742 (6.4)	1.07 (0.98, 1.17)	1.13 (1.03, 1.24)
Complicated appendicitis
Not complicated	2849/69593 (4.1)	Referent	Referent
Complicated	2878/26349 (10.9)	2.81 (2.67, 2.96)	2.77 (2.63, 2.93)
Sex
Female	2328/38038 (6.1)	1.04 (0.99, 1.10)	1.03 (0.98, 1.09)
Male	3399/57904 (5.9)	Referent	Referent

Figure 1.

Kaplan-Meier curves depicting time to late hospital care for variables with a significant association.

For both 30- and 180-day follow-up intervals, children with late hospital care had higher costs, which were largely due to ED and inpatient expenses (Table 3). Among children with no late hospital care, costs were driven primarily by the initial encounter while for those with late hospital care, costs were primarily due to expenditures from later encounters.

Table 3:

Median and IQR costs for patients who did and did not have late hospital care. Costs are given for all care within 30 or 180 days. For determination of 30-day costs, late complications must have occurred within 30 days; thus we analyzed 64,852 (67.6%) for the 30-day analysis. For determination of 180-day costs, they must have occurred within 180 days; thus we analyzed 47,137 (49.1%) for the 180-day analysis.

	No late hospital care, median $ (IQR)	Late hospital care, median $ (IQR)	p
30-day costs (n=64,852)	4925 (1465, 13762)	17026 (7340, 49932)	<0.001
Initial encounter	4317 (774, 12893)	5330 (348, 16378)	0.52
Later encounters	245 (142, 545)	10226 (3487, 26645)	<0.001
ED/hospital	0 (0, 0)	9401 (2795, 24694)	<0.001
Outpatient	228 (134, 450)	317 (136, 1007)	<0.001
180-day costs (n=48,816)	6553 (3021, 16171)	19589 (8642, 64384)	<0.001
Initial encounter	4292 (792, 12812)	5460 (609, 16925)	0.001
Later encounters	1563 (981, 2687)	12520 (4691, 37418)	<0.001
ED/hospital	0 (0, 0)	9186 (2085, 29616)	<0.001
Outpatient	1448 (919, 2354)	1897 (1093, 4296)	<0.001

Discussion

Among nearly 100,000 children with appendicitis enrolled in Medicaid, rehospitalization or subsequent surgery was uncommon, occurring in 4.4% of children within 30 days and 6.0% within 180 days. Children < 5 years old, those with CCCs, complicated appendicitis, and children living in higher opportunity areas had a lower likelihood of late hospital care. Late hospital care as associated with substantial increases in costs, driven primarily by ED and inpatient care and not by additional outpatient utilization. Together, our findings indicate that late hospital care is uncommon but are important and costly.

The reasons for late hospital care are unsurprising. Complicated appendicitis leads to worse outcomes including need for multiple surgeries, sepsis, gut injury, and adhesions from abdominal abscess formation, and thus would be expected to lead to rehospitalization or late surgeries.¹¹ CCCs make the care of most conditions more challenging, predispose to hospitalizations, and create higher opportunities for surgical complications.¹² The youngest children are at higher risk for delayed diagnosis of appendicitis leading to complicated appendicitis.^13,14 While we adjusted for complicated appendicitis status, there are likely residual differences in severity of presentation by age group that account for higher rates of late complications in children.

Late hospital care was associated with higher costs. While this was an expected finding, the extent of increased costs was high. In the cohort with 180-day follow-up, initial costs were also higher, suggesting that those who go on to have late hospital care have more complicated early courses, which is further evidenced by their higher complicated appendicitis rates and longer index length of stay.

This work has several implications. First, reducing the need for late hospital care is likely to yield substantial cost savings. Second, prevention efforts should focus on improving care in the youngest patients, medically complex patients, and those with complicated appendicitis. Prevention of complicated appendicitis could be particularly useful, which may be improved by timely diagnosis since timely diagnosis is associated with lower complicated appendicitis rates.¹⁵ Finally, our work has importance for future research; research that does consider care occurring beyond the initial encounter will underestimate the rate of complications, overall costs and burden to patients and families. Given that 75% of late hospital care occurred within 33 days following the ED visit or hospitalization during which appendicitis was diagnosed, and given that seeming unrelated disorders more commonly happened outside the 30-day window (e.g. depressive disorders), we believe a 30-day follow-up period should be used for outcomes-based research in pediatric appendicitis.

There were several limitations. First, we treated same-day or following-day claims as transfers and therefore may have underestimated the proportion of late hospital care if a patient had an immediate return visit. Second, although diagnosis codes are generally accurate for appendicitis (correct in >95% of cases in a previous study), they are not perfect, and are subject to miscategorization.⁹ We believe that would be infrequent and unlikely to affect conclusions. Finally, patients on certain capitated Medicaid plans did not have expenditure data, and their expenditures likely differ from those on fee for service plans or where the claims processor shares fee for service equivalents.

In conclusion, late hospital care for pediatric appendicitis is infrequent but does occur and should be measured in appendicitis outcomes research. Reducing the need for late hospital care is likely to reduce the overall cost of appendicitis care.