Impact of Low Birth Weight on the Prevalence and Economic Burden of Common Childhood Illnesses Among Under-Five Children in India: Findings From Nationally Representative Surveys

Background Common childhood illnesses such as diarrhea, fever, and acute respiratory infection impose substantial health burdens among under-five children, and Low Birth Weight (LBW) has been associated with an increased prevalence of these illnesses. However, the impact of LBW on healthcare utilization and the economic burden of these illnesses remains understudied. Aim To assess the impact of LBW on the prevalence, healthcare utilization, and Out of Pocket Expenditure (OOPE) for outpatient (OP) treatment of selected Common Childhood Illnesses (CCHI) among under-five children in India. Methodology This study utilized data from two nationally representative surveys conducted in India; National Family Health Survey (NFHS-5) (2019-2021) and the National Sample Survey Organization (NSSO) 75th Round Schedule Social Consumption: Health (2017-2018). Data from the NFHS-5 was analyzed to assess the impact of LBW on the prevalence of selected CCHI and healthcare utilization. Comparison of OOPE for OP treatment of selected CCHI between LBW and Normal Birth Weight (NBW) children done using the median OOPE for OP visits of CCHI estimated from the NSSO data. Results The two-week prevalence of selected CCHI among LBW and NBW children was found to be 20.0% (95% CI 19.6 -20.4) and 18.0% (95% CI 17.8 -18.2), respectively. There was no significant difference between LBW and NBW children on healthcare utilization for the treatment of CCHI; both groups had a similar proportion (around 70%) of formal medical treatment utilization for CCHI. The median OOPE spending for OP visits per episode of CCHI was comparable between LBW and NBW children. However, families of LBW children had higher annual OOPE spending for OP visits related to CCHI, with projected estimates of INR 1,446 ($19.56) for LBW children and INR 1,271 ($17.2) for NBW children. Conclusion LBW was associated with a higher prevalence of CCHI. Even though healthcare utilization was similar among LBW and NBW children, a higher prevalence of CCHI among LBW children led to higher OOPE. LBW children have approximately 13% higher annual OOPE spending for the OP visits related to selected CCHI compared to NBW children.


Introduction
Low birth weight (LBW) is a crucial public health indicator that can provide insights into various factors related to child health, nutrition, healthcare delivery, and poverty [1]. It is an important measure for assessing the survival and overall health of newborns [2]. Since birth weight is strongly associated with the health and nutritional status of the mother, it closely reflects the health status of the communities into which they are born [3].
The World Health Organization (WHO) defines LBW as a birth weight of less than 2,500 grams regardless of gestational week at delivery [4]. The estimated global prevalence of LBW in 2015 was 14.6%, with the highest prevalence (26.4%) reported from Southern Asia [5]. A more recent report, the National Family Health Survey (NFHS-5) (2019-2021), indicates that the prevalence of LBW in India is 18.2% [6].
LBW can challenge the health of children in a number of ways; children born with LBW had a higher risk of mortality and morbidity, malnutrition in childhood, certain child malignancies, developmental delays, etc. [7][8][9][10][11]. LBW is also associated with long-term physical ill health in adulthood, which includes cardiovascular disease, renal diseases, psychiatric disorders, cardiometabolic diseases, etc. [12][13].
Common childhood illnesses such as fever, acute respiratory infections (ARI), and diarrhoea are the major contributors to the under-five mortality rate [14]. The 2020 global disease burden study revealed that lower respiratory tract infections and diarrhoea are the second and third leading causes of Disability-Adjusted Life Years (DALYs) among children aged 0 to 9 years in 2019 [15]. A cohort study done in India from 2014 to 2016 found that respiratory tract infections, gastrointestinal illness, and fever are the highest reported illness among children during the first two years and accounted for approximately 74% of all reported morbidity episodes [16].
Several studies conducted globally have found that LBW is associated with a higher incidence of childhood infections [11,[17][18][19]. National Health Account Estimates in India for 2019-20 reported that 52% of healthcare expenditures are met through out-of-pocket spending [20]. A study on data analysis of the National Sample Survey Organization (NSSO) 75th schedule reported that the OOPE for outpatient (OP) care and hospitalization in households reporting childhood infections were 4.8% and 6.7% of their total consumption expenditure in India [21]. Childhood infections lead to poor health outcomes and impose a significant financial burden on households and society [21]. Since LBW is strongly associated with a higher incidence of childhood illnesses and long-term health issues, which can have significant financial implications for households and healthcare systems, particularly in low-and middle-income countries.
Although many studies worldwide have shown a strong association between LBW and childhood morbidity [11,[17][18], there has been limited research exploring the financial implications associated with it. Many of these studies were conducted as small-scale cohort studies. In this study, we aim to assess the impact of LBW on the prevalence of selected common childhood illnesses, healthcare utilization, and OOPE spending of this illness by analyzing data from two nationally representative surveys such as NFHS-5 and NSSO 75th Schedule for Social Consumption Health. Our research has the potential to provide valuable insights into the influence of LBW on the prevalence of CCHI, healthcare utilization, and the additional financial burden placed on households due to the management of these illnesses for LBW children.

Sources of data
This study utilized data from two nationally representative surveys conducted in India.

NFHS-5 Survey (2019-2021)
The NFHS-5 survey encompassed 636,699 households, 724,115 women, and 101,839 men from all states and Union Territories (UT) of India. The survey utilized a two-stage, stratified cluster sampling design and covered topics on fertility, mortality, family planning, maternal and child health, reproductive health, nutrition, anemia, and the utilization and quality of healthcare services, etc. The detailed NFHS -5 survey methodology was provided elsewhere [6].

NSSO 75th Round for Schedule Social Consumption: Health (2017-18)
The NSSO 75th schedule survey covered 1,13,823 households and 5,55,115 persons from all the states and U.T.s from India. A stratified multi-stage design was adopted for this survey. The survey collected data on health indicators like disease prevalence, healthcare utilization, expenses, and socio-economic characteristics such as income, education, and occupation. The detailed NSSO 75th schedule methodology was published elsewhere [22].

Study variables and operational definitions
We have included children who have completed the age range of 0 to 4 years from both surveys as under-five children. Infants with a birth weight below 2,500g are classified as LBW. To ensure data accuracy, infants with missing or implausible birth weight measurements (birth weight less than 500g or more than 6,000g) were excluded from the study. As the NSSO survey did not capture the birth weight of children, birth weight classification was only possible in NFHS-5 data.
Both surveys collected data on the two-week prevalence of common childhood illnesses and healthcare utilization. The NFHS-5 survey captured data on the occurrence of three major common childhood illnesses such as fever, diarrhoea, and symptoms of respiratory illness such as cough, shortness of breath, and chest problems. In NFHS-5, participants were asked to recall whether the child had experienced any of these illnesses during the last two weeks and indicate their responses as "Yes" (present), "No" (not present), or "Don't Know" (unknown). Whereas the NSSO survey recorded the occurrence of various ailments during the last 15 days recall period by classifying them into 60 different health conditions without specific responses for each morbidity. Both NFHS-5 and NSSO followed a similar operational definition for diarrhoea which includes the passage of three or more liquid stools per day. The operational definitions for ARI were slightly different in both surveys, NFHS-5 operational definitions of ARI included the presence of short, rapid breathing or difficult breathing which were chest-related whereas NSSO definitions of ARI included the presence of running nose, cough, and sore throat, with or without fever all of the short duration, though it could be recurrent or cough as the main symptom, with or without fever, with or without sputum and blood in it, with or without marked breathlessness. To harmonize the operational definitions of fever between the two surveys, we have categorized fever without any accompanying symptoms of respiratory illness in the NFHS-5 survey as "Fever". For this study, the presence of any of these three morbidities (diarrhoea/ARI/fever) is classified as Common Childhood Illness (CCHI).
Both the NFHS-5 and NSSO surveys provided information on healthcare utilization for CCHI. The healthcare utilization patterns for CCHI were evaluated based on the type of treatment received, categorized as no treatment, informal treatment, treatment in public facilities, and treatment in private facilities. In the NSSO survey dataset, informal treatment is directly recorded and defined as treatment provided by untrained health service providers working outside regulatory frameworks. On the other hand, in the NFHS-5 survey, responses for informal treatment were derived from a set of variables, which included treatment from pharmacies, paramedical personnel, shops, friends or relatives, home remedies, and other non-formal treatments. NSSO survey provided direct responses for treatment taken in public health facilities which included all public institutions such as sub-centers, primary health centers (PHCs), community health centers (CHCs), etc. In NFHS 5 treatment from the public facility was derived from a set of variables such as government hospitals, government AYUSH, PHCs, CHCs, sub-centers, government mobile clinics, public camps, and anganawadi. The NSSO survey provided a direct response for treatment taken in private facilities which included private doctors or clinics and private hospitals. In the NFHS-5 survey treatment from the private facility was derived from a set of variables which included treatment from a private hospital/clinic, private doctor, private AYUSH, and other private institutions. In this study, since the number of individuals receiving treatment from NGOs or charitable institutes was low in both surveys, these facilities have been included in the private facility category for the purpose of analysis. An outcome variable, a formal medical treatment created in both surveys, refers to treatment sought in either private or public facilities.
NSSO survey provided details of OOPE for CCHI whereas NFHS-5 did not collect information on OOPE for CCHI. We defined Direct OOPE as spending on doctor/surgeon fees, medicines, diagnostic services, and other medical expenses and Indirect OOPE as expenditure on transportation fees, expenditure for food, etc.

Covariates
The NFHS-5 data set was used to assess the other factors related to the prevalence of CCHI and healthcare utilization. These included the sex of the child (male/female), age of the children was expressed in completed months and classified into four groups (less than six months, 6 to 12 months, 13 to 36 months, and 37 to 59 months). The birth order of the child was categorized as one, two to three, or more than three. Place of delivery was grouped into three categories: home delivery, delivery in a government facility, and delivery in a private facility. The type of delivery was classified as either caesarean or vaginal. The place of residence was categorized as urban or rural. The social group included categories such as schedule caste, schedule tribe, other backward castes, and others. Religion was classified as Hindu, Muslim, Christian, and others. The age of the mother was expressed in completed years and grouped into four categories: 15 to 20 years, 21 to 29 years, 30 to 39 years, and 40 years and above. The education of the mother was categorized as no education, primary education (grades 1 to 5), secondary education (grades 6 to 12), and higher education (grades 12 and above). The wealth index, a composite measure calculated from household assets and durable possessions, was divided into quintiles ranging from 1 (poorest) to 5 (richest) [6]. Indian States and Union Territories (UT) grouped based on the Social Progress Index (SPI) which evaluated the progress of states and districts based on 12 components related to basic human needs, foundations of well-being, and opportunities. Based on the SPI index States/UTs were classified into high, middle, and low social progress categories [23]. Covariates were selected based on the factors associated with morbidity and related expenditures known from the literature and available in the datasets.

Statistical analyses
All the statistical analyses were done using STATA version 14.0 (Stata Corp LP, College Station, TX, USA). Sample weights provided with the NFHS and NSSO data set were used for descriptive analysis. All the percentages reported in the study are weighted percentages. To account for the complex survey design, the 'svy" suite command was used in the NFHS -5 analysis for the Pearson Chi-square test to assess the difference in the prevalence of CCHI and healthcare utilization for CCHI with selected LBW covariates. Unweighted multiple binary logistic regression was used to assess LBW's association with the prevalence of CCHI and formal medical treatment. The Median OOPE estimates, calculated from NSSO data, were adjusted for inflation to 2021 price using the Consumer Price Index (CPI) and these estimates were further converted to United States dollars (US $) using the average annual exchange rate of 2021 (1 US $=₹73.92).
The OOPE for an average episode of illness and annual average OOPE spending for CCHI among LBW and NBW children was calculated by the following formulas. To project the two-week OOPE estimates to annual OOPE estimates we have applied a multiplication factor of 26.1 (52 weeks/2). The annual projected OOPE was calculated by assuming the same patterns of prevalence of CCHI and health expenditure throughout the year.
To apply this formula, we used median adjusted OOPE estimates from NSSO data and health care utilization percentage for CCHI, the number of children with CCHI, and the total number of children from the NFHS-5 data. Figure 1 shows the selection of under-five children from NFHS-5 survey data for analyzing the impact of LBW on the prevalence of CCHI and treatment-seeking patterns. About 7.5% of children were not weighed at birth, and for nearly 2% of children, birth weight was not known to mothers. The percentage of under-five children with Extremely Low Birth Weight (ELBW) and Very Low Birth Weight (VLBW) was 0.06% (n =106) and 0.95% (n=1,683) respectively. A total of 202,817 children were included in the analysis. The Mean (SD) birth weight of selected LBW and NBW children were 2007g (312) and 2980g (428) respectively.     Table 3 compares the healthcare utilization pattern of LBW and NBW children for CCHI. In both groups medical treatment was sought for nearly 70% of CCHI. The highest medical treatment was sought for fever (74%) and the least for ARI (51.6%). We observed that many socio-economic background characteristics were highly associated with the utilization of formal medical treatment ( Table 4).

Discussion
In this study, we have assessed the impact of LBW on prevalence, healthcare utilization, and OOPE for the OP treatment of CCHI in India. This study utilized data from two large-scale national representative surveys; the NSSO 75th Schedule of social consumption health and the NFHS-5 survey.
Based on our analysis of the National Family Health Survey-5 data, we have observed a higher prevalence of all CCHI in children with LBW compared to those with NBW. Despite the relatively short recall period (2 weeks), the large sample size of the NFHS-5 survey provides robust estimates of the association between LBW and CCHI. Our findings align with several prior studies that have also reported an increased prevalence of CCHI in LBW children [11,[17][18]. These results indicate that LBW constitutes a significant risk factor for the development of CCHI in children, emphasizing the necessity for targeted interventions to enhance the health outcomes of this vulnerable population.
Our analysis of the NFHS-5 data revealed that the proportion of formal medical treatment utilization for CCHI was 70%. However, this estimate is lower than the analysis from the NSSO survey, which reported a higher utilization of formal medical treatment for CCHI (79%). This difference in estimates could be attributed to various factors, such as differences in the survey methodologies, sample sizes, and population characteristics. However, despite these differences, the results of these surveys can be considered comparable as both surveys covered a large representative sample of households in India and are similar to results from other smaller studies in various parts of India. A cohort study conducted in the rural area of Pune, India, in 2015-16 reported that formal treatment was sought for 70% of episodes of illness among under-five children [24]. Similarly, a cross-sectional study of tribal children from south India also reported that 70% of under-five children sought formal medical treatment [25]. A higher prevalence of seeking formal treatment (84%) was observed from a cross-sectional study done in the urban slums of Orissa, India in 2016-17 [26]. These differences in the prevalence of formal medical treatment can be attributed to differences in settings and various other cultural and health system factors.
Results of this study show no difference in the healthcare utilization between LBW and NBW children; although a good sign, does not reflect the anxiety of families having children with LBW, which would lead to higher healthcare utilization. On the contrary, a cohort study from Ghana in 2013 reported a higher proportion of absence of care seeking among LBW children compared with NBW children [27]. This is expected given the factors leading to LBW also contribute to poor health-seeking behaviors.
We estimated the median OOPE for OP treatment of selected CCHI in India using the NSSO 75th schedule dataset. While there is a published study that analyzed the same dataset to assess the average OOPE per episode of outpatient visit and hospitalization in children aged less than five years for selected infectious diseases [21], they reported the average OOPE for outpatient visits and hospitalization for common childhood illnesses. We have conducted a different analysis using the same dataset to estimate the median OOPE for outpatient visits of three selected common childhood illnesses. In our study, we have found that the median OOPE for OP treatment of diarrhoea, fever, and ARI was 470₹, 328₹, and 410₹ respectively. We also estimated the median OOPE for OP treatment of all three selected CCHI together as 400₹. Similar to our finding a study from slums of eastern India in 2019 reported the median OOPE for outpatient visits of childhood illness as 326₹ [28] and another study from urban Puducherry, on the southeast coast of India in 2015, reported the median OOPE for childhood illness as 375₹ [29].
In our study, we aimed to compare the OOPE for OP treatment between LBW and NBW children. To achieve this, we applied the adjusted median OOPE of CCHI derived from our analysis of the NSSO dataset to the NFHS results on the prevalence and healthcare utilization patterns of CCHI. Despite high prevalence rates of CCHI among LBW children, we did not find any significant difference in healthcare utilization patterns between the two groups. Therefore, the average OOPE for outpatient visits per episode of illness was similar for both groups. However, due to the higher prevalence of CCHI among LBW children, the total OOPE incurred for this group was greater than that of NBW children. A cohort study from Mozambique found significantly higher household spending for health care costs of LBW infants compared to NBW infants. A report of the NSSO 75th schedule survey on health found that over half of the households (52 percent) faced catastrophic health expenditure when seeking treatment for their one-to-two-year-old children at private hospitals [30].
The present study contributes to the limited literature on the comparison of the OOPE for CCHI between LBW and NBW children in India by providing comparative estimates for outpatient visits for selected CCHI. The use of two nationally representative surveys enhances the generalizability of our findings to the larger Indian population. We have also taken care of the inflation while applying the OOPE estimates from the 2017-18 survey to the 2019-21 survey by calculating the expenditures for 2019-2 with the use of a multiplication factor based on the Consumer Price Indices for these years. However, there are some limitations to our study. Firstly, as a secondary data analysis, we were able to analyze only three selected CCHI, and the estimates may not be representative of the overall burden of childhood illnesses in India, which is likely to be higher increasing the difference estimated between the two groups. Additionally, the prevalence estimates were based on a shorter duration, which could also be viewed as an advantage in reducing the recall bias. However, the other major shortcomings may be the variations in the operational definitions of illness and healthcare utilization of these two surveys. Furthermore, as we only analyzed outpatient visits, our findings do not provide information on the cost of inpatient visits, which are typically more expensive than outpatient visits. If included, this would only inflate the differences observed in the OOPE between the LBW and NBW children.

Conclusions
Our study showed that LBW children had a higher prevalence of CCHI compared with NBW children. We have also observed higher OOPE spending for OP treatment related to CCHI among LBW children than NBW children. Although our study focused on a limited duration of two weeks and OP visits, the impact of LBW on CCHI prevalence was still evident. Despite numerous studies examining the relationship between LBW and morbidities, very few have explored the financial burden of LBW. Our findings emphasize the need for further research on the financial implications of LBW at the individual and household levels.

Additional Information Disclosures
Human subjects: Consent was obtained or waived by all participants in this study. Animal subjects: All authors have confirmed that this study did not involve animal subjects or tissue. Conflicts of interest: In compliance with the ICMJE uniform disclosure form, all authors declare the following: Payment/services info: All authors have declared that no financial support was received from any organization for the submitted work. Financial relationships: All authors have declared that they have no financial relationships at present or within the previous three years with any organizations that might have an interest in the submitted work. Other relationships: All authors have declared that there are no other relationships or activities that could appear to have influenced the submitted work.