Abstract
Background Data on COVID-19 clinical characteristics and severity from resource-limited settings are limited. This study examined clinical characteristics and factors associated with COVID-19 mortality and hospitalisation in rural settings of Indonesia, from 1 January to 31 July, 2021.
Methods This retrospective cohort included individuals diagnosed with COVID-19 based on polymerase chain reaction or rapid antigen diagnostic test, from Lampung, Gorontalo, Central Sulawesi, Southeast Sulawesi, and East Nusa Tenggara Provinces. We extracted demographic and clinical data, including hospitalisation and mortality from COVID-19 surveillance records. We used mixed-effect logistic regression to examine factors associated with COVID-19-related mortality and hospitalisation.
Results Of 6,583 confirmed cases, 205 (3.1%) died, and 1,727 (26%) were hospitalised. The median age was 37 years (IQR 26-52), with 825 (12·53%) under 20 years, and 3,371 (51.21%) females. 4,533 (68.86%) cases were symptomatic, 319 (4.85%) had a clinical diagnosis of pneumonia, and 945 (14.36%) with at least one pre-existing comorbidity. The mortality and hospitalisation rate ranged from 2.0% and 13.4% in East Nusa Tenggara to 4.3% and 36·1% in Lampung. Age-specific mortality rates were 0.9% (2/340) for 0-4 years; 0% (0/112) for 5-9 years; 0.2% (1/498) for 10-19 years; 0.8% (11/1,385) for 20-29 years; 0.9% (12/1,382) for 30-39 years; 2% (23/1,095) for 40-49 years; 5% (57/1,064) for 50-59 years; 11% (62/576) for 60-69 years; 16% (37/232) for ≥70 years. Older age, pre-existing diabetes, liver diseases, malignancy, and pneumonia were associated with higher risk of mortality and hospitalisation. Pre-existing hypertension, cardiac diseases, chronic kidney disease, COPD, and immunocompromised condition were associated with risk of hospitalisation but not with mortality.
Conclusion Clinical characteristics and risk factors of severe COVID-19 outcomes in rural provinces were broadly similar to those in urban settings. The risk of COVID-19-related mortality and hospitalisation was associated with higher age, pre-existing chronic comorbidities, and clinical presentation of pneumonia.
Background
The coronavirus disease 2019 (COVID-19) pandemic continues to affect the world population. Globally, there have been 537,591,764 total confirmed cases, and 6,319,395 deaths reported to the World Health Organization (WHO) as of 21 June 2022 [1]. As per 20 June 2022, almost 12 billion COVID-19 vaccine doses have been administered worldwide [1], yet 39,717 new COVID-19-related deaths were still being reported in the last 28 days [2].
Current evidence suggests that severe outcomes of COVID-19 have been consistently associated with older age and pre-existing chronic conditions, such as hypertension, diabetes, obesity, cardiac disease, chronic kidney disease, and liver disease [3–8]. Moreover, recent findings from US [11-12], Chile [13], Brazil [14], and including from Indonesia [15] have suggested that poor outcomes associated with COVID-19 are concentrated in groups with higher socio-demographic and health care vulnerability. However, the understanding of COVID-19 severity mostly comes from studies conducted in urban settings. Limited data exist from rural settings on COVID-19 clinical characteristics and factors associated with severity where differences in age distribution, comorbidities, access to quality health services, and other factors may influence trends regarding severe outcomes [8,15-18].
Indonesia is a low-middle income country (LMIC) that has suffered the highest number of COVID-19 confirmed cases and deaths in Southeast Asia, second only to India in Asia [2]. Since the first reported cases on 2 March 2020, Indonesia has reported a total of 6,070,933 cases and 156,700 deaths (2·6% confirmed case fatality rate) up to 21 June 2022 [19]. The majority of cases and deaths in Indonesia were reported in Java Island, a more developed setting populated by 152 million individuals (56% of total Indonesia’s population). Recent studies from Indonesia’s capital city of Jakarta suggested that COVID-19 disproportionately affected individuals with older age and pre-existing chronic comorbidities, as well as those areas with lower vaccine coverage and higher poverty and population density [8,15]. However, data on the impact of COVID-19 in less developed areas of Indonesia with greater geographic, cultural and socio-economic diversity, outside of the island of Java, are scarce.
The limited evidence regarding COVID-19 epidemiology and severity from low resource settings in Indonesia – especially in areas outside of Java - has been particularly associated with the lack of epidemiological data recorded by the surveillance systems in these areas. To improve quality of COVID-19 surveillance and response in Indonesia, the Indonesian Epidemiologist Association (Perhimpunan Ahli Epidemiologi Indonesia, PAEI) in collaboration with the Ministry of Health of Indonesia implemented a new COVID-19 information system named Sistem Informasi Surveilans Epidemiologi (SISUGI) [20] in five rural provinces – East Nusa Tenggara in eastern Indonesia, Gorontalo, Central Sulawesi, and Southeast Sulawesi on the island of Sulawesi, and Lampung on the island of Sumatra (Figure 1). This study utilised the epidemiological surveillance data recorded in the SISUGI, to better understand clinical characteristics and factors associated with COVID-19-related mortality and hospitalisation in five rural provinces in Indonesia, from January to July 2021.
The administrative boundaries of the study sites were highlighted in bold. COVID-19 aggregated data were obtained from https://data.covid19.go.id/.
Methods
Study design and participants
This was a retrospective cohort study to assess demographic and clinical factors associated with mortality and hospitalisation of adults and children diagnosed with COVID-19 in five provinces in Indonesia. The study population included individuals diagnosed with COVID-19 based on either a polymerase chain reaction (PCR) test or rapid antigen diagnostic test (Ag RDT) from Lampung, Gorontalo, Central Sulawesi, Southeast Sulawesi, and East Nusa Tenggara provinces between January to July 2021. In accordance with Indonesia’s national COVID-19 guidelines, individuals are categorized as COVID-19 confirmed cases if they tested positive by Ag RDT or PCR [21]. Based on data from Indonesia Central Bureau of Statistics, the majority of areas in the five provinces were categorised as rural areas i.e., Lampung (68%), Gorontalo (66%), Central Sulawesi (55%), Southeast Sulawesi (83%), and East Nusa Tenggara (92%) [22, 23]. Based on the number of doctors, nurses, midwives, and public health officer per 100,000 population (Supplementary Table 1), these five provinces were considered as rural settings with limited health care capacity. The five provinces were purposively selected as pilot project sites for the implementation SISUGI (https://sisugi.com/) developed by PAEI in collaboration with the US CDC Indonesia Country Office, South Asia Field Epidemiology and Technology Network (SAFETYNET), and Ministry of Health of the Republic of Indonesia. This study was conducted as part of a program evaluation.
Patient and public involvement
This study did not include patients and public in the design, or conduct, or reporting, or dissemination plans. This study was a secondary analysis of routine surveillance data conducted under a formal collaboration between Indonesian Epidemiologist Association, the US CDC Indonesia Country Office, and Ministry of Health of the Republic of Indonesia.
Data collection
Each health facility had a surveillance officer responsible for conducting an epidemiological investigation of each confirmed case via interview using a standard form. Completed forms were submitted to the SISUGI database. Data regarding dates of test, onset of illness, SARS-CoV-2 Ag RDT and PCR testing, and outcomes (mortality and hospital admission) were recorded, along with age, gender, symptoms, and comorbidities. Comorbidities were recorded based on clinical assessments at admission or patient reporting. Fever was defined as axillary temperature of ≥ 38°C. Clinical diagnosis of pneumonia (indicative of moderate or severe disease [24]) was determined by the treating physician, based on clinical and radiological evaluation [25]. Individual-level epidemiological data of all COVID-19 confirmed cases between January to July 2021 were then extracted from the SISUGI database. Province-level data were collected from various government sources. Data on population numbers in 2020 were collected from the Statistics Bureau Database [26]. Data on the number of doctors, nurses, midwives, as per October 2021 were collected from the Indonesia Ministry of Health records.
Statistical analysis
Descriptive statistics included proportions for categorical variables and medians and interquartile ranges (IQRs) for continuous variables. We used the Mann-Whitney U test, Chi-squared test, or Fisher’s exact test to compare characteristics between deceased and recovered cases, and between hospitalised and non-hospitalised cases. Due to the high proportion of missing data needed to define the time from onset to admission, this variable was not analysed.
We used bivariable and multivariable mixed effects logistic regression models to determine the risk of death and hospitalisation, expressed as odds ratio with 95% confidence intervals. Province was treated as the random effect variable to adjust for clustering of observations within provinces. We did null model analysis (no predictor was added) and the result justified the use of the mixed effects models. All independent variables with p-value <0.10 in bivariable analysis were included in the multivariable models. Final model selection was informed by likelihood ratio tests. We reported two final multivariable logistic regression models. In the first model, we assessed the effect of each type of comorbidity (hypertension, diabetes, cardiac disease, chronic obstructive pulmonary disease (COPD), chronic kidney disease, immunocompromised status, and liver disease), and in the second model, the number of comorbidities (0, 1 or >1), along with other demographic and clinical factors. We used interaction terms to examine potential effect modification by age and sex. We set statistical significance at 0·05, and all tests were two sided. All analyses were done in Stata/IC 15.1 (StataCorp, College Station, TX, USA). This study is reported as per Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines [27].
Ethics
This study was approved by the Health Research Ethics Committee of Universitas Negeri Semarang (300/2/KEPK/EC/2022). The requirement for patient consent was waived as this was a secondary analysis of anonymised routine surveillance data.
Role of the funding source
The funder of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report. The corresponding author had full access to all of the data and the final responsibility to submit for publication.
Results
General characteristics of study participants
Between 1 January and 31 July 2021, a total of 6,583 confirmed COVID-19 cases were recorded on the SISUGI. The completeness of data was high, with less than 1% missing data across the variables analysed in the study. The study flow chart and completeness of data are presented in Figure 2.
Table 1 presents the demographic and clinical characteristics of the 6,583 cases included in the analysis, as well as health care worker numbers in the studied provinces. The highest proportion of cases were reported in Lampung (32.60%, 2,146/6,583), and the lowest was in East Nusa Tenggara (10.0%, 658/6,583). The median age was 37 years (IQR 26-52, range 0·1-158), with 825 (12·53%) under 20 years, and 3,371 (51.21%) were females. The overall median number of health care workers was 6.57 (IQR: 4.97-8.47) doctors per 100,000 population, 17.41 (IQR: 10.52-27.46) nurses per 100,000 population, 13.03 (IQR: 12.74-19.18) midwives per 100,000 population, and 47.60 (IQR: 9.20-51.40) public health officers per 100,000 population.
Clinical characteristics and COVID-19-related mortality and hospitalisation
Of 6,583 cases included in the analysis, 205 (3.1%) were deceased, and 1,727 (26·2%) were hospitalised. The number of COVID-19-related deaths and hospital admissions varied over time and by province (Supplementary Figure 1). There were 4,533 (68.86%) symptomatic cases, with the most common presenting symptom being cough (46.67%, 3,072), followed by fever (43.96%, 2,894), runny nose (33.02%, 2,174), malaise (25.38%, 1,671), and others (Figure 3A). The median number of symptoms was 2 (IQR 0–4, range 1–13). There were 319 (4.85%) patients with a clinical diagnosis of pneumonia. Overall, 14.36% (945) of cases had a record of one or more pre-existing comorbidities, including hypertension (8.19%, 539), diabetes mellitus (5.54%, 365), cardiac disease (2.84%, 187), COPD (0.73%, 48), and others (Figure 3B).
Compared to recovered cases, deceased cases were older (median 58 vs 36 years); more likely to have more presenting symptoms (median 3 vs 2); to have cough, malaise, myalgia, shortness of breath, nausea/vomiting, abdominal pain, and diarrhoea (Table 1); to have pneumonia (33% vs 4%); and to have a history of any comorbidity (56% vs 12%), 1 comorbidity (34% vs 10%) or >1 comorbidities (26% vs 3%), specifically hypertension (33% vs 7%), diabetes (29% vs 5%), cardiac disease (15% vs 03%), COPD (5% vs 0.6%), chronic kidney disease (8% vs 0.5%), malignancy (2% vs 0.3%), and liver disease (2% vs 0·2%) (Table 1).
Compared to non-hospitalised cases, hospitalised cases were older (median 50 vs 34 years); more likely to have more presenting symptoms (median 3 vs 1); to have cough, fever, malaise, headache, myalgia, shortness of breath, nausea/vomiting, abdominal pain, and diarrhoea (Table 1); to have pneumonia (16% vs 1%); and to have a history of any comorbidity (32% vs 6%), 1 comorbidity (25% vs 6%) or >1 comorbidities (10% vs 1%), specifically hypertension (19% vs 4%), diabetes (14% vs 2%), cardiac disease (2% vs 0.2%), chronic kidney disease (2% vs 0.22%), malignancy (0.6% vs 0.2%), immunocompromised (0.6% vs 0.1%), and liver disease (0.6% vs 0·1%) (Table 1).
Heterogeneity of COVID-19-related mortality and hospitalisation rate
The mortality and hospitalisation rate increased with age (Table 2). Whilst age-specific mortality rate ranged from 0.9% (2/340) for 0-4 years to 16% (37/232) for ≥70 years, the age-specific hospitalisation rate ranged from 17% (37/215) for 0-4 years to 57% (131/232) for ≥70 years. Based on province, the mortality and hospitalisation rate ranged from 2% (13/658) and 13% (88/658) in East Nusa Tenggara to 4% (93/2,146) and 36% (775/2,146) in Lampung. Details on age-specific mortality and hospitalisation rate can be seen in Table 2.
Factors associated with COVID-19-related mortality and hospitalisation
In bivariable mixed effects logistic regression analysis (Supplementary Table 2), the risk of death was associated with older age, clinical diagnosis with pneumonia, number of pre-existing comorbidities, hypertension, diabetes, cardiac disease, COPD, chronic kidney disease, liver diseases, malignancy, lower number of nurses per 100,000 population, and higher number of midwives per 100,000 population. The risk of being hospitalised was associated with older age, clinical diagnosis with pneumonia, number of pre-existing comorbidities, hypertension, diabetes, cardiac disease, COPD, chronic kidney disease, liver diseases, malignancy, immunocompromised, and lower number of nurses per 100,000 populations.
Based on the first multivariable mixed effects logistic regression model (Table 3), the risk of death increased for age groups 50-59 years (aOR 6·47, 95%CI 1·96-21·33), 60-69 years (aOR 10·76, 95%CI 3·23-35·80), and ≥70 years (aOR 18·32, 95%CI 5·31-63·23), compared to <20 years; for cases with pre-existing diabetes (aOR 2·02, 95%CI 1·29-3·15), chronic kidney disease (aOR 6·05, 95%CI 2·52-14·55), liver diseases (aOR 10·34, 95% CI 1·82-58·68), or malignancy (aOR 7·37, 95% CI 1·48-36·86); for patients who had pneumonia (aOR 8·00, 95%CI 5·34-11·98).
The risk of being hospitalised increased for age groups 40-49 years (aOR 1·67, 95%CI 1·27-2·18), 50-59 years (aOR 2·90, 95%CI 2·22-3·79), 60-69 years (aOR 3·26, 95%CI 2·40-4·42), and ≥70 years (aOR 3·70, 95%CI 2·50-5·49), compared to <20 years; for cases with pre-existing hypertension (aOR 2·36, 95%CI 1·89-2·98), diabetes (aOR 2·19, 95%CI 1·65-2·90), cardiac disease (aOR 2·08, 95% CI 1·40-3·09), COPD (aOR 7·46, 95% CI 2·32-24·03), chronic kidney disease (aOR 3·58, 95%CI 1·57-8·15), liver diseases (aOR 7·45, 95% CI 1·79-31·00), malignancy (aOR 3·46, 95% CI 1·14-10·45), or immunocompromised (aOR 6·93, 95% CI 2·38-20·19); for patients who had pneumonia (aOR 15·32, 95%CI 10·74-21·84).
In the second multivariable model (Supplementary Table 3), we found that both the risk of death and hospitalisation increased with the number of pre-existing comorbidities, after controlling for age, sex, pneumonia, and number of health care workers. Sex and number of health care workers at province-level were not associated with COVID-19-related hospitalisation and mortality. Sex was not found to be an effect modifier in either model.
Discussion
This retrospective cohort study described the clinical characteristics and factors associated with risk of COVID-19-related mortality and hospitalisation based on epidemiological surveillance data of confirmed COVID-19 cases reported from five rural provinces of Indonesia, during January to July 2021. This analysis was one of the first COVID-19 epidemiological studies reporting data from rural settings of Indonesia. The observed disease pattern broadly reflects that reported globally, with patients usually presenting multiple symptoms of fever, cough, malaise and/or shortness of breath. The overall mortality and hospitalisation rate was 3% (205/6,583) and 26% (1,727/6,583), respectively. The age-specific mortality and hospitalisation rate for children under five years old was 0.9% (2/340) and 17% (37/215), respectively. The mortality and hospitalisation rate were highly heterogeneous, with the highest rate was reported in Lampung (4% mortality and 36% hospitalisation), and the lowest rate in East Nusa Tenggara (2% mortality and 13% hospitalisation). Higher risk of mortality and hospital admission was associated with increased age, clinical diagnosis with pneumonia, presence of pre-existing diabetes, chronic kidney disease, liver disease, and malignancy. Presence of pre-existing hypertension, cardiac diseases, COPD, and immunocompromised conditions were associated with higher risk of hospitalisation, but not associated with mortality. Gender and number of health care workers at province-level were not associated with COVID-19-related mortality and hospitalisation.
The overall mortality rate in these five rural provinces was higher than the mortality rate reported from Jakarta (3% vs 1.5%), potentially due to a higher prevalence of at least one comorbidity in these settings (14% vs 0.7%). However, our findings suggest that the hospitalisation rate was slightly lower than in Jakarta (26% vs 33%), possibly due to a lower level of transmission, a lower health care capacity, and a lower access to health care facility in these five provinces. For comparison, as presented in Supplementary Table 1, the highest capacity in these settings was seen in Southeast Sulawesi, with 13 doctors per 100,000 population, 37 nurses per 100,000 population, and 13 midwives per 100,000 population. By contrast, Jakarta has 45 doctors per 100,000 population, 48 nurses per 100,000 population, and 31 midwives per 100,000 population. In addition, as previously reported in Jakarta [15], the inequality of COVID-19 vaccine coverage may also contribute to the mortality and hospitalisation rate in Jakarta, but this was not evaluated in the present study. Further study assessing association between health care capacity, including vaccine coverage, with COVID-19-related mortality and hospitalisation is needed to better understand Indonesia’s health system resilience in responding the current and future global health crisis.
Consistent with a recent finding from an urban area of Indonesia, Jakarta [15], the mortality rate for children under five years old in these rural settings was very low (<1%). This finding was also consistent with several other studies from China [25], Brazil [29,30], Uganda [16], and South Africa [17] that have reported COVID-19-related deaths among children under 5 years to be rare. Moreover, a review of data from US, Korea and Europe in the early phase of epidemic estimated the overall case fatality rate among children (0–19 years old) to be as low as 0.03 per 100 000 (44/42 846) [31].
Our findings suggest that COVID-19-related mortality and hospitalisation rates varied across the five provinces. For example, compared to in East Nusa Tenggara, the rate in Lampung was 2 times (4% vs 2%) for mortality and 2.8 times (36% vs 13%) for hospitalisation. The observed different risks of COVID-19-related mortality and hospitalisation could be due to several factors such as different prevalence of older and comorbid population, as well as different level of health care capacity across the provinces. Although our study found an absence of association between number of health care workers at province-level with COVID-19-related mortality and hospitalisation, a previous nationwide ecological study in Indonesia, suggested that COVID-19 has disproportionately affected district with high proportion of elder population, high prevalence of diabetes mellitus, lower doctor to population ratio, higher life expectancy at birth, and lower level of formal education [32]. Other health care factors such as quality of COVID-19 diagnosis and case management may also affect the COVID-19 heterogeneity, but these factors were not evaluated in the current study.
Risk of mortality and hospitalisation increased with the presence of one or more comorbidities, especially pre-existing diabetes, chronic kidney diseases, liver diseases, or malignancy, and clinical pneumonia. The mortality risk increased by 102%, 505%, 934%, 637%, and 700% for people with pre-existing diabetes, chronic kidney diseases, liver diseases, malignancy, and a clinical diagnosis with pneumonia, respectively. The presence of pre-existing hypertension, cardiac diseases, COPD, or immunocompromised conditions also increased risk of hospitalisation by 136%, 108%, 646%, and 593%, respectively, but was not associated with increased risk of death. In addition, mortality and hospitalisation risk were increased for people which were clinically diagnosed with pneumonia. These findings are consistent with evidence from previous studies suggesting that pre-existing comorbidities, including cardiovascular or cerebrovascular diseases, hypertension, and diabetes mellitus, were associated with poorer COVID-19 outcomes [5–7, 33]. Although chronic comorbidities among COVID-19 cases in these five rural provinces (14%) were less frequent than reported in more developed settings including North America [5,6], South America [23], and Europe [7], this finding could reflect the under-reporting of comorbidities by patients, or under-diagnosis due to variable access to quality health services. Finally, obesity has also been recognised as an important risk factor of mortality [7,18], but we could not assess this due to absence of data for this variable.
This study has limitations. The retrospective design and reliance on routine surveillance data meant that data were incomplete or unavailable for some key baseline variables (e.g., date of symptom onset, date of hospital admission and outcome, vital signs, and disease severity classification). The imperfect contact tracing, testing, and reporting activities could result in underreporting of cases, especially for asymptomatic and mild cases which thus results in overestimation of rates of mortality and hospitalisation rate. Comorbidities were often self-reported or could be under-diagnosed, potentially resulting in underreporting and hence underestimation of effect sizes. Findings from this study may not reflect the mortality and hospitalisation rate, and risk factors associated with COVID-19-related outcomes in other LMIC settings with different demographics, non-infectious diseases prevalence, and different health care capacity.
In conclusion, clinical characteristics and risk factors associated with severe COVID-19 outcomes in five rural provinces in Indonesia are broadly similar to those in the urban province of Jakarta, and in more developed countries in North America, Europe, and Asia, dominated by advanced age and comorbidities. Heterogeneity of mortality and hospitalisation rate in the studied areas was likely driven by the different age structure and prevalence of pre-existing comorbidities. The findings highlight the need for enhanced context-specific public health action to reduce mortality and hospitalisation risk among older and comorbid rural populations, and improved care and treatment for the cases. Although health care worker capacity was not associated with mortality and hospitalisation in this study, future nationwide study assessing association between health care capacity and COVID-19 impacts is needed to better guide improvement of health systems resilience in the future.
Data Availability
After publication, the datasets used for this study will be made available to others on reasonable requests to the corresponding author, including a detailed research proposal, study objectives and statistical analysis plan.
Contributors
HS was the principal investigator of this study. HS designed the study, did the analysis, and had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. AME, MR, WN, JADH contributed to data collection and verification. ENM, YYS, SS, RR, IM, and HW supervised the data collection and verification. CYP, WAH, and HW contributed to funding acquisition for the SISUGI development and implementation. HS, LF and DR obtained the ethical approval. HS, CYP, IRFE, and WAH drafted the paper. All authors critically revised the manuscript for important intellectual content and all authors gave final approval for the version to be published.
Conflict of interests
We declare no competing interests.
Data sharing statement
After publication, the datasets used for this study will be made available to others on reasonable requests to the corresponding author, including a detailed research proposal, study objectives and statistical analysis plan.
Funding
This study received no specific grant from any funding agency in the public, commercial or not-for-profit sectors. HS and IRFE were funded by the Wellcome (UK) Africa Asia Programme Vietnam (106680/Z/14/Z).
Acknowledgments
We acknowledge South Asia Field Epidemiology and Technology Network (SAFETYNET) for their support in the development and implementation of SISUGI, and all health care workers involved in the care for the COVID-19 patients, as well as those involved in the field data collection.