Clinical characteristics and outcomes in patients with diabetes mellitus affected by COVID-19: a retrospective cross-sectional study from a tertiary care center in Pakistan

Abstract

Background

COVID-19, caused by SARS-CoV-2, emerged in December 2019 and quickly became a global public health concern. Diabetes, a major risk factor for severe COVID-19, affects 537 million people worldwide, with high prevalence in low- and middle-income countries like Pakistan. Studies show diabetes increases the risk of severe COVID-19 complications and mortality. However, there is limited data on COVID-19 outcomes in diabetic patients in Pakistan. This study aims to fill this gap and examine factors affecting outcomes in this population.

Methods

We conducted a retrospective cross-sectional study at The Aga Khan University Hospital, Karachi, encompassing 2,346 confirmed COVID-19 patients from February 26, 2020, to September 6, 2021. Data on diabetic status, following ADA guidelines and other clinical outcomes were collected from medical records and patient interviews. Statistical analysis was performed using SPSS V.25.

Results

A total of 1,342 patients were included, with 864 males (64.4%) and 478 females (35.6%). The mean age was 56.59 ± 15.55 years. SARS-CoV-2 infection was the primary diagnosis for 741 patients (55.2%), while 601 patients (44.8%) had it as a secondary diagnosis. Of the total, 348 patients (25.9%) had T2DM, 2 patients (0.15%) had T1DM, and 991 patients (73.8%) were non-diabetic. The mean duration of diabetes was 2.01 ± 1.32 years. Diabetic patients had a significantly shorter mean hospital stay (4.99 ± 4.46 days) compared to non-diabetic patients (6.79 ± 7.32 days) (p < 0.001). The overall discharge rate was 70.3%, with a mortality rate of 10.7%. T2DM was associated with lower in-hospital mortality (p < 0.001) but higher rates of ARDS (p < 0.001). There was no significant association between T2DM and the risk of pulmonary aspergillosis, pulmonary embolism, or septic shock. Higher financial class was associated with longer hospital stays and a greater likelihood of being discharged home (p < 0.001).

Conclusion

In conclusion, our study highlights the heightened susceptibility of COVID-19 patients with concurrent T1DM and T2DM to developing ARDS. Despite no significant association found between diabetes and adverse outcomes, the crucial role of tailored care for high-risk groups, particularly those with diabetes, cannot be overstated.

Background

The coronavirus outbreak (COVID-19), caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), emerged as a global threat in Wuhan, China at the end of December 2019 [1]. By the end of 2020, the virus had affected over 40.68 million individuals across 218 countries, resulting in more than 1.2 million deaths worldwide [2].

COVID-19 exhibits a highly variable presentation, ranging from asymptomatic to acute respiratory failure [3]. Consequently, COVID-19 patients have been classified into four categories: mild, moderate, severe, and critical cases [4]. Several risk factors for severe COVID-19 and adverse outcomes have been identified, with underlying diabetes mellitus being one of the most well-established [5]. Diabetes mellitus comprises of a group of metabolic disorders characterized by hyperglycemia due to defects in insulin secretion, insulin action, or both [6]. As of 2021, diabetes imposes a significant global burden, affecting 537 million people worldwide, with 3 out of 4 diabetic individuals residing in low- or low-and-middle-income countries (LMICs) [7]. South-Asian countries are known to have some of the highest prevalence of diabetes globally. Notably, Pakistan, a South Asian LMIC, ranked third globally in 2021 with an alarming 33 million cases [8].

In recent years, multiple studies have investigated the effect of diabetes with SARS-CoV-2 susceptibility, complications and outcomes, revealing greater severity and mortality in diabetic patients compared to those without diabetes [9–13]. This is particularly important as diabetes is one of the most frequent comorbidities presenting in COVID-19 patients, with a global prevalence reaching up to 14.7% [14]. Various meta-analyses have also explored the impact of diabetes on COVID-19 patients, reporting that diabetic patients have a two-fold increase in both severity and mortality compared to non-diabetic patients [15]. Additionally, another meta-analysis revealed that diabetes is associated with a higher risk of developing complications, particularly acute respiratory distress syndrome (ARDS), leading to poorer disease progression in these patients and higher rates of mortality [16].

Mild to moderate elevations in glucose levels, regardless of a prior history of diabetes, have also been widely reported in patients suffering from COVID-19. The exact pathophysiology of hyperglycemia in COVID-19 remains unknown but is believed to be due to multiple factors, including stress-induced beta-cell dysfunction, steroid-induced hyperglycemia, and an inflammatory state caused by cytokine storms in severe ARDS secondary to COVID-19. Patients with uncontrolled hyperglycemia experienced longer hospital stays and higher mortality rates [17–19]. Additionally, the risks associated with uncontrolled hyperglycemia are heightened in patients with pre-existing diabetes, obesity, and other risk factors [20].

Due to these elevations in glucose levels being seen, studies have concluded that the possibility of COVID-19 causing new-onset diabetes and having lasting impacts on the secretion and effects of insulin cannot be excluded [18, 20]. Therefore, a bidirectional relationship has been established between COVID-19 and diabetes, with ineffective phagocytic activity, reduced T-cell function, and diminished innate and adaptive immunity making diabetics more susceptible to developing COVID-19; whereas decreased insulin secretion and increased resistance lead to a higher risk of developing diabetes in COVID-19 patients. Additionally, recent evidence also indicates that SARS-CoV-2 can cause direct damage to the pancreas, making patients more susceptible to the induction of new-onset diabetes in previously healthy subjects [21]. Similarly, a study from the UK reported 30 children with new-onset T1DM, representing an 80% increase in new-onset T1D during the pandemic compared with previous years [22]. This provides strong evidence in support of COVID-19-inducing diabetes in previously non-diabetic patients. Furthermore, the risk of mortality appears to be higher in people with new-onset diabetes than in COVID-19 patients with a previously recorded history of diabetes [18].

There is limited data on the prognosis of COVID-19 infections in diabetic patients within the Pakistani population, leaving a critical gap in understanding the interplay between these two significant health challenges. This study seeks to address this gap by exploring the prevalence of diabetes among patients with COVID-19 and identifying factors that influence outcomes in this demographic. The mechanisms underlying new-onset diabetes related to COVID-19, as well as its epidemiology and optimal management, remain incompletely understood. Investigating these issues in a country like Pakistan, where Type 2 diabetes mellitus (T2DM) is already highly prevalent, can provide valuable insights into the unique challenges faced by this population.

Methods

Study design

We conducted a retrospective cross-sectional study in order to evaluate the prevalence of diabetes in patients with COVID-19 and analyze the differences in baseline characteristics and their respective clinical courses.

Setting and participants

The sample included 2346 patients with confirmed cases of COVID-19. The diagnosis of COVID-19 was confirmed based on a positive viral real-time reverse transcriptase-polymerase chain reaction (RT-PCR) assay. We included all COVID-19-positive patients admitted at Aga Khan University Hospital (AKUH), between the 26th of February 2020 till the 6th of September 2021 by reviewing the medical records from our hospital. AKUH is a large tertiary care hospital located in Karachi, the most populous city in Pakistan, and was one of the first centers to offer care to patients with COVID-19. Diabetic patients were identified based on the patient’s medical records or self-report.

Data collection

All patients aged 18 or less were excluded from the study. Data was extracted after the approval of the ethical review committee. The data was collected through a unique identification number. Medical history, medications, routine laboratory results, length of stay, and other clinical outcomes were collected from the electronic medical record. Demographic details, which included socioeconomic class, education, occupation, etc., were collected after approaching the patients using their contact details. Patients’ confidentiality was protected with no direct patient identifiers used in data collection and entry.

Operational definitions

Diabetes was defined in accordance to the American Diabetes Association (ADA), where diabetes in nonpregnant individuals was diagnosed based on one or more of the following criteria: an HbA1C ≥ 6.5%, fasting plasma glucose (FPG) ≥ 126 mg/dL after at least 8 h of fasting, or plasma glucose ≥ 200 mg/dL two hours after a 75 g oral glucose tolerance test (OGTT). In individuals with classic hyperglycemia symptoms or a hyperglycemic crisis, random plasma glucose ≥ 200 mg/dL was considered to be diagnostic [23].

The higher financial class was defined as households with an annual income greater than PKR 3 million and two dependents at most.

Primary outcomes

The primary outcome of the study was to assess the prevalence of diabetes in patients with COVID-19 and the differences between the baseline characteristics including the demographic details, duration and type of diabetes, clinical characteristics, comorbid conditions, length of stay and complications during the stay along with recommended treatment and guidance after discharge of patients who suffered from Diabetes and COVID-19.

Outcomes and statistical analysis

Patients were divided into two groups on the basis of their current diabetic status. In order to compare outcomes between the study groups, a statistical analysis was done via the Statistical Package for Social Sciences (SPSS) V.25 (IBM, New York, USA). Categorical variables were reported as frequencies with proportions and continuous variables as means with standard deviations (SDs). Prevalence crude and adjusted ratios with 95% Confidence Interval (CI) were reported for univariate and multivariable analysis.

Results

The total number of participants included in the study was 1342, with 864 (64.4%) being male and 478 (35.6%) being female. The mean age of patients was 56.589 (SD = 15.55). Figure 1 gives more detail about the inclusion and outcomes of these patients.

Fig. 1.

shows a flow diagram of patients included and their outcomes in the study

SARS-COVID-19 infection was the primary diagnosis for 741 (55.2%) patients, whilst being a secondary diagnosis for the remaining 601 (44.8%) patients. These patients were seeking care for a non-COVID-related illness but were diagnosed with COVID-19 at presentation or at any point during their hospital stay. The majority of the patients were admitted via ER (85.7%), followed by admission requests from consulting outpatient clinics (9.46%).

348 (25.9%) of the patients had a diagnosis of T2DM, 2 (0.15%) had T1DM and 1 patient had gestational diabetes, whereas 991 (73.8%) were non-diabetic. 245 patients had concurrent hypertension along with diabetes, 84 had renal dysfunction (60 with chronic kidney disease), and 79 had ischemic heart disease (IHD) alongside diabetes. The mean duration since diagnosis of diabetes amongst patients was 2.0078 years (SD = 1.3205).

The mean in-hospital length of stay (LOS) for all patients was 6.32 days (SD = 6.73). The reported length of stay for diabetic patients was significantly lower (4.99, SD = 4.46) as compared to non-diabetic patients (6.79, SD = 7.32) (p < 0.05). Please refer to Table 1 for more details. Of all patients treated at our center, 944 (70.3%) were eventually discharged from the hospital, 143 (10.7%) expired, 128 (9.54%) were discharged on request, and 127 (9.46%) left against medical advice (LAMA).

Table 1.

Shows the dichotomized length of stay in the hospital for both diabetic and non-diabetic COVID-19 patients

LOS (days)	Diabetics n (%)	Non-Diabetics n (%)	Total n (%)
0	7 (0.52%)	27 (2.01%)	34 (2.53)
1–2	87 (6.48%)	253 (18.9%)	340 (25.3)
3–4	114 (8.49%)	224 (16.7%)	338 (25.2)
5–6	68 (5.07%)	140 (10.4%)	208 (15.5)
7–8	33 (2.46%)	100 (7.45%)	133 (9.91)
8+	41 (3.06%)	248 (18.5%)	289 (21.5)

A diagnosis of T2DM was associated with lower in-hospital mortality (p < 0.05) from COVID-19 infection when compared to non-diabetics. Please refer to Table 2 for details about in-hospital mortality in diabetic and non-diabetic COVID-19 patients. Higher rates of ARDS secondary to COVID-19 (p < 0.05) were observed in T2DM patients, however, there was no significant association between T2DM and the risk of developing pulmonary aspergillosis, pulmonary embolism, or septic shock in COVID-19 infection. The financial class of the patients had a significant association with the discharge deposition and length of stay, with patients of higher financial class staying longer and being more likely to be sent home than LAMA or expiring (p < 0.05).

Table 2.

Shows in-hospital mortality for both diabetic and non-diabetic COVID-19 patients

Expired	Diabetics n (%)	Non-Diabetics n (%)	Total n (%)
Yes	17 (1.27%)	126 (9.39%)	143 (10.7)
No	333 (24.8%)	866 (64.5%)	1199 (89.3)

Discussion

Examining the interplay between diabetes and COVID-19, the findings of this study unveil notable trends. Emphasizing the coexistence of diabetes with comorbidities such as hypertension, chronic kidney disease, and cardiac issues, the study indicates unexpected phenomena, including truncated hospital stays and diminished in-hospital mortality rates among diabetic patients.

The intertwining nature of hypertension and diabetes is evident, as highlighted by a high incidence of concurrent hypertension among diabetic patients (18.2%). A retrospective study in Wuhan, China, identified hypertension as the most common comorbidity in COVID-19 patients with diabetes [24].

The elevated prevalence of concurrent health conditions, including hypertension, chronic kidney disease, cardiac disease, and cerebrovascular disease, in diabetic patients underscores the necessity for a comprehensive healthcare strategy. Within our study cohort, renal dysfunction was found in 84 patients, with 60 displaying chronic kidney disease, and 79 presenting with a past history of ischemic heart disease (IHD). There were 245 patients who had concurrent hypertension alongside diabetes. This aligns with a significantly increased occurrence of these comorbidities in diabetic patients compared to their non-diabetic counterparts, consistent with findings from a retrospective 7-month follow-up cohort study in Tehran, Iran, in 2020, and further supported by a 2022 case-control study in Jamshoro, Pakistan, examining clinical outcomes in diabetic versus non-diabetic COVID-19 patients [25, 26]. This increased prevalence of multiple health issues in diabetic COVID-19 patients likely contributed to their admission and influenced subsequent outcomes throughout the course of the illness.

Contrary to expectations, diabetic patients exhibit a reduced length of hospital stay and lower in-hospital mortality compared to their non-diabetic counterparts. This unexpected trend may be attributed to the shorter LOS, and lower in-hospital mortality in diabetic patients may be attributed to various factors highlighted in prior literature, including distinct admission criteria, variations in disease severity prompting admission, diabetes type considerations, and the age distribution of patients [15]. Previous data also supports the mortality rate in critically ill patients to be significantly higher in patients without a history of diabetes mellitus compared to those with diabetes [15, 27, 28].

The decreased in-hospital mortality among diabetic patients can be attributed to more efficient and targeted care, established protocols, regular consultations, and monitoring for associated health issues. Diabetic patients and their caregivers may also adhere more closely to treatment plans due to their awareness of long-term health risks. Advances in diabetic management and outpatient care also play a role in decreasing the necessity for prolonged hospital stays. This rationale has been supported by previous studies [27]. Consequently, this may lead to a quicker recovery and shorter hospitalization period, further emphasizing the potential reduction in the need for prolonged hospital stays with advancements in diabetic management and outpatient care. It is also possible that only healthier diabetic patients (with fewer severe comorbidities) have the opportunity to be admitted, whereas more critically ill patients might be less likely to reach tertiary care centers, skewing the observed outcomes.

A 2021 study examining the association between hyperglycemia and COVID-19 severity and outcomes, regardless of a diagnosis of pre-existing diabetes revealed a significant increase in the impact of hyperglycemia on mortality among non-diabetic patients, challenging the typical focus on glycemic control primarily in diabetic individuals. These findings show that patients with diabetes may exhibit a potential resilience to acute hyperglycemia during critical illnesses like COVID-19, underscoring the need for attention to glycemic management in all inpatients. In addition, the study hints at the possibility of more aggressive hyperglycemia management in diabetic patients due to heightened awareness among both patients and healthcare providers, which explains the shorter length of hospital stay as well as lower mortality [29].

While this study did not find an increased mortality rate in diabetic patients, we did find that diabetes significantly increases the risk of acute respiratory distress syndrome (p < 0.05). This finding is consistent with multiple other studies (19,20,22). A study by Bradley et al. found that patients with diabetes had a higher risk of developing ARDS after COVID-19 infection [30]. Another study by Joshua et al. also found that metabolic syndrome significantly increased the odds of ARDS following COVID-19 [31]. The reason behind this could be that type 2 diabetes mellitus (T2DM) with hyperglycemia is among the factors that lead to elevated expression of angiotensin-converting enzyme 2 (ACE2) in lungs and other tissues; ACE2 is the cellular “receptor” and port of viral entry. The existing chronic inflammation, coupled with an intensified inflammatory response to the infection and rising viral levels, can trigger a severe systemic immune reaction known as a “cytokine storm,” which is strongly associated with the heightened severity of COVID-19 [31].

The findings indicated that patients belonging to a higher financial class tended to have prolonged hospital stays [32]. Several factors may contribute to this association. Individuals with higher financial status may opt for extended hospital durations to facilitate thorough health monitoring. In contrast, those with a lower financial status may be influenced by financial considerations in their decision-making process regarding healthcare seeking and the duration of their hospital stay. Furthermore, the higher financial class might exhibit an overutilization of healthcare resources, given their ability to access and request more medical interventions or services. In addition, higher financial status individuals are more likely to be sent home compared to lower financial status individuals, who predominantly leave against medical advice or expire. These findings highlight the economic burden on families in low- and middle-income countries (LMIC) and highlight how financial considerations significantly influence healthcare decisions, sometimes leading to premature hospital departures.

This study includes patient data from one tertiary care private hospital in a metropolitan city in Pakistan. Therefore, the results (including LOS, mortality, etc.) for a COVID-19-related admission, might not be representative of national figures since an overwhelming majority of patients still seek healthcare in free public and smaller primary and secondary care hospitals. Another limitation of the study was the exclusion of nearly 40% (1004) of patient records due to incomplete or missing data. Due to the retrospective nature of our study, the available evidence does not allow for a conclusive determination of a temporal relationship between diabetes and COVID-19.

Conclusion

Our study confirms that COVID-19 patients with T1DM and T2DM have an increased risk of developing ARDS, consistent with previous literature. However, in contrast to prior evidence, we found no association between diabetes and in-hospital mortality or length of stay—in fact, the opposite. These findings may be influenced by study design and single-center limitations. Further longitudinal studies are needed to assess the long-term impact of COVID-19 in diabetic patients and optimize multidisciplinary management strategies.

Author contributions

S.A., and A.A. worked towards the conception of the idea and development of the research protocol. All authors worked towards data collection. S.A. and M.A. worked on the data analysis. All authors contributed to the manuscript.

Funding

No funding was provided by any institutions or individuals for this study.

Data availability

All data generated or analyzed during this study are included in this published article.

Declarations

Ethical approval and consent to participate

Ethical approval for this study was obtained from the Ethics Review Committee at Aga Khan University (ERC Reference No: 2021-6820-19956). As this was a retrospective study utilizing pre-existing data without direct patient interaction, the Ethics Review Committee granted a waiver of informed consent in accordance with ERC guidelines, such that informed consent was not required.

Clinical trial number

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.