Scaling-up skilled health professionals in the rural Indian public health system: projections for 2030 Authors: Yash Jawale, Siddhesh Zadey, Varad Puntambekar, Vedika Sharma, Vidhi Wadhwani, Rachit Sekhrajka, Amit Sahu, Nanditha Venkatesan, Aatmika Nair, Suraj Bhor, Sweta Dubey, Surabhi Dharmadhikari Published: 2026 ABSTRACT BackgroundIndia faces a shortage of skilled health professionals (SHPs). There is limited research measuring SHP deficits in rural public health centres. We estimated the current and future SHP densities and deficits, along with costs for SHP scale-up to achieve Sustainable Development Goal (SDG) 2030 targets.MethodsWe used the number of SHPs (doctors, nurses, and midwives) at rural primary and community health centres from the Rural Health Statistics reports (2009–2019) to calculate average annual percentage changes (AAPCs) at national and state levels using JoinPoint regression. Using AAPC values, we projected SHP counts and estimated densities (per 10,000 people) for years until 2030. The projected deficits for 2030 were calculated for three target density thresholds, aligning with the Millennium and Sustainable Development Goals. For scale-up costs, a state-wise statistical average of salaries for SHP groups was calculated assuming a 5% annual increase in allowance.ResultsDuring 2009-19, SHP density grew annually by 2.59% (95% CI: 0.93–4.28). The national SHP density would increase from 8.85 SHPs per 10,000 in 2019 to 11.47 in 2030. In 2030, India is expected to have a deficit of 0.47–1.83 million SHPs. The scale-up costs to cover these deficits would range from INR 1.46–4.96 trillion, about 15% of India’s government health spending.ConclusionAt the current rate, Indian rural public health centres will not achieve the target SHP density by 2030. The costs for scaling up the recruitment and retention of rural SHPs are relatively small, making it feasible.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12913-026-14550-x. Background Greater availability of skilled health professionals (SHPs), i.e., doctors, nurses, and midwives, is associated with greater universal health coverage [1]. Existing literature reports that India’s SHP density is considerably lower than the recommended 44.5 SHPs per 10,000 people necessary to achieve the Sustainable Development Goals (SDGs) and 22.8 for the Millennium Development Goals (MDGs) [2–5]. Previous assessment using National Sample Survey (NSS) 2017-18 data reported a national density of 16.7 SHPs per 10,000 people, which dropped to 11 when considering only well-qualified professionals [6]. Beyond availability, the maldistribution of SHPs across public vs. private healthcare sectors and rural vs. urban regions, and states is a concern. Over 80% of doctors and 70% of nurses work in the private sector, catering to less than 50% of the patients [7]. Two-thirds of Indians reside in rural regions. However, they are served by less than one-third of SHPs [6]. The proportion of doctors practising in rural India went down from 34.5% in 2011 to 27.4% in 2017, and their densities have decreased from 3.04 per 10,000 to 2.4 per 10,000; while that of nurses marginally increased from 33.0% in 2011 to 35.7% in 2017, however, their density has decreased from 3.36 per 10,000 in 2001 to 3.10 in 2016 [6, 7]. Limited access to SHPs can compel people in rural regions to travel long distances often, miss out on a day’s wages, and incur significant out-of-pocket expenditures on healthcare and transportation costs, even for most primary and secondary healthcare services [5, 6]. While the current analysis focuses primarily on the shortage, it is important to note that beyond the availability of SHPs, their knowledge, skills, and acceptability to the local population, and supporting infrastructure (e.g., diagnostics/lab services, essential drugs, etc.) determine the capacity and quality of healthcare provided. Consequently, rural populations lag behind their urban counterparts on several health outcomes, including preventable mortality [8]. To cater to the basic healthcare needs of large rural populations, the Government of India launched the National Rural Health Mission (NRHM) in 2005, subsequently subsumed under the National Health Mission (NHM) [9]. The mission aimed to expand primary and secondary-level care by establishing and upgrading multiple centres along the referral hierarchy. These centres have operational guidelines– Indian Public Health Standards (IPHS), ensuring adequate infrastructural and human resources [10]. Previously, we noted that the IPHS thresholds might not accurately capture the true adequacy of health workforce distribution [11]. Adequacy can be better defined based on established population thresholds aligned with global literature and aspirational health policy targets. The World Health Organization (WHO) noted a minimum density threshold of 22.8 SHPs per 10,000 people to attain 80% coverage for skilled birth attendance in its 2006 report [12]. In its 2016 report, WHO further noted a threshold of 44.5 SHPs per 10,000 people to attain 80% coverage on twelve health indicators selected to reflect SDG targets [13]. In line with these, the National Institute for Transforming India (NITI) Aayog, Government of India (GoI) set an ambitious target of 55 SHPs per 10,000 people for the public sector in its 2018-19 report [14]. Hence, we adapted these three target SHP density thresholds as: Acute (skilled birth attendance proxy), WHO-SDG, and India-SDG at the rural level to achieve 22.8, 44.5, and 55 SHPs per 10,000 people, respectively. There is a paucity of research measuring SHP density in rural areas, particularly in their public health centres, relative to these thresholds. There is also no evidence of progress over the years, a potential chance of target achievement by 2030, or financing required to achieve the targets. To fill this gap, we investigated the current state and future needs for SHPs in the rural public health system at primary and secondary levels in India. We had three aims. First, to understand the trends of SHP densities and deficits for primary and community health centres (PHCs and CHCs) at the national and subnational levels from 2009 to 2019. Second, to estimate the projected SHP density for 2030, current vs. desired annual growth rates, and the number of additional SHPs needed to achieve three target thresholds. Finally, estimate deployment and retention costs required for the SHP scale-up to inform pragmatic health policy-making and fund allocation. Methods SHP density and deficit calculations We extracted national and state-level annual ‘in-position’ (present staff) data for SHP cadres: allopathic doctors such as physicians, surgeons, paediatricians, obstetricians and gynaecologists (OBGYNs), general duty medical officers (GDMOs - MBBS doctors), and nurses at rural community health centres (CHCs); and the number of GDMOs, nurses, and auxiliary nurse midwives (ANMs) at rural primary health centres (PHCs) and ANMs at sub-centres (SCs) for 2009–2019 from the Rural Health Statistics (RHS) reports [15, 16]. We did not include AYUSH practitioners (Ayurveda, Yoga, Naturopathy, Unani, Siddha, and Homeopathy) or Community Health Officers (CHOs) in our SHP calculations, as consistent data on AYUSH practitioners was unavailable for several years, and CHOs were only recently introduced in 2018. We address this exclusion as a limitation with only a minor effect on SHP density in the discussion section. We obtained mid-year annual rural population projections from the relevant Population Council reports [17, 18]. We analyzed data for India and its 35 states and Union Territories (UTs). We did not include Telangana separately in this analysis as it was considered under Andhra Pradesh until 2014, which is a midpoint for the study period from 2009 to 2019. The public sector service utilization proportion (SUP) was captured using the proportion of hospitalizations for all broad aliment types at government/public hospitals, including the PHCs and CHCs, among rural residents using the data from the 75th Round of the NSS on Social Consumption in Health (2017–2018) [19]. We considered two scenarios for estimating SHP densities and deficits. We estimated SHP densities (per 10,000 people) based on total rural and SUP-adjusted rural populations, termed as aspirational and realistic densities, respectively (Eq. 1). For realistic estimation, we calculated the SUP-adjusted rural population by multiplying the rural population by SUP (Eq. 2). We used three thresholds (per 10,000 people) − 22.8 for the Acute [12], 44.5 for the WHO-SDG [13], and 55 for the India-SDG target [14] to calculate the required number of SHPs by multiplying each threshold with total and SUP-adjusted rural populations separately (Eq. 3). We calculated SHP deficits as the absolute difference in required and available SHP for all three thresholds (Acute, WHO-SDG, and India-SDG), assuming two scenarios: aspirational and realistic (SUP-adjusted) (Eq. 4).1\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\textrm{SHP Density}\:}_{i}=\frac{\textrm{Available SHP}}{{\textrm{Rural Population}\:}_{i}}\times 10000$$\end{document}2\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\textrm{Population}\:}_{\textrm{realistic}}={\textrm{Population}\:}_{\textrm{aspiration}}\times{\textrm{SUP}\:}_{\textrm{fraction}}$$\end{document}3\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\textrm{Required SHP}\:}_{i,j}={\textrm{Threshold}\:}_{j}\times {\textrm{Rural Population}\:}_{i}$$\end{document}4\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\textrm{SHP Deficit}\:}_{i,j} = {\textrm{Required SHP}\:}_{i,j} - \textrm{Available SHP}$$\end{document} Where, i = aspirational or realistic densities. j = Acute, WHO-SDG, India-SDG thresholds. AAPC calculation Average annual percentage change (AAPC) was a measure of the overall growth rate for a variable over time (i.e., overall slope) (Eq. 5). We fitted the data with a minimum number of consecutive linear fits of different slopes (i.e., piece-wise linear regression), where each slope depicted an annual percentage change (APC) or annual growth rate. The intersection point of consecutive linear fits was a joinpoint. The AAPC was the time-period weighted average of the consecutive corresponding linear fits. In other words, AAPC was the year-wise weighted average of individual APCs separated by joinpoints [20]. 5\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\textrm{AAPC}=\frac{{\sum\:}_{k=1}({w}_{k}\times{\textrm{APC}\:}_{k})}{{\sum\:}_{k=1}{w}_{k}}$$\end{document} Where, w = period of linear fit. k = linear fit. The joinpoint model used log-transformed SHP count data with autocorrelated errors and a constant variance homoscedasticity assumption. For parsimony, the model was capped to find a maximum of two joinpoints, given that we only had data for eleven years. We selected the best-fitted model with the optimal number of changepoints based on a permutation test. We used 4499 permutations with a 0.05 significance level using the default parameters of the permutation test in the Joinpoint. The model also provided a 95% confidence interval (CI) for AAPC values. We calculated AAPCs for SHPs from 2009 to 2019 for India and 35 states/UTs with Joinpoint (Command-Line (Batch/Callable) version 4.8.0.1) [21]. We used data from the years till the most recent year for the states with complete data from 2009 to 2019. For Arunachal Pradesh, Delhi, Jharkhand, Karnataka, Maharashtra, Mizoram, Odisha, and Sikkim, the 2019 data were missing, while for Bihar and Chandigarh, 2018 and 2019 were missing. We excluded these datapoints for the joinpoint analyses of individual states. Overall analysis was performed using Python Jupyter Notebooks (Python 3.7.3, Anaconda, Inc.). Projected densities, deficits, and AAPC targets We projected the number of SHPs for 2030 using the 2019 base data (or from the most recent year for which complete data was available) based on the AAPC values (Eq. 6). Assuming SHPs grow at the rate of AAPC each year, leading to compounding growth over time. Thus, an exponential growth rate formulation provides direct projections for the target year.6\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${N}_{(l+n)}= {N}_{l}\:(1+r{)}^{n}$$\end{document} Where, N = number of SHPs for the lth year (initial year) and l + nth year (target year). n = number of years projected (i.e., n ~ (l + n) - l). r = AAPC or exponential growth rate. We calculated SHP densities in 2030 (with Eq. 1) using the total and SUP-adjusted rural population projections for aspirational and realistic scenarios, respectively. The required number of SHPs in 2030 was also reverse calculated using Eq. 1, with the three thresholds – Acute, WHO-SDG, and India-SDG, as required SHP densities. We calculated SHP deficits as the absolute difference in the required and projected numbers of SHPs. The 95% CIs for projections and deficits were calculated using the 95% CIs for AAPC values. We calculated the AAPCs required to achieve the three threshold SHP densities by 2030 by solving the exponential growth rate equation for the target rate (r’) (Eq. 7).7\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$r^{\prime}={e}^{\left(\frac{log\left(\frac{{D}_{y}}{{D}_{x}}\right)}{y-x}\right)}-1$$\end{document} Where, r’ = Required or target AAPC. Dx = SHP densities in 2019. Dy = Target thresholds. y-x = Period of 11 years (i.e., 2030 − 2019). Separately, we also calculated the time required to achieve different thresholds by solving the exponential growth rate equation for a period (n’) with 2019 SHP densities as initial, corresponding target thresholds, and the current AAPC as the growth rate (r) (Eq. 8). The projected densities beyond 2030 from Eq. 8 do not account for population changes in the subsequent (post-2030) years.8\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$n^{\prime}=\frac{log\left(\frac{{D}_{y}}{{D}_{x}}\right)}{log(1+r)}\:;\:y\leftrightarrow x+n$$\end{document} Where, n’ = year when the target density will be achieved at the current AAPC. r = AAPC (i.e., current growth rate). Dx = SHP densities in 2019 (base data). Dy = Target thresholds. Projection validation We extracted state-level annual ‘in-position’ SHP data (as mentioned in Sect. “SHP density and deficit calculations”) from RHS 2020 and 2021 for validation [19]. SHP counts projected for 2020 and 2021 using AAPC were compared with actual numbers of SHPs using Lin’s Concordance Correlation Coefficient (CCC) [22]. The agreement was categorized as almost perfect for CCC > 0.99, substantial for CCC [0.95–0.99], moderate for CCC [0.90–0.95], and poor for CCC < 0.90. The confidence interval (CI) of CCC was calculated using the Z-transformation approach [23]. Cost estimations We estimated the costs – money to be invested by the governments – to overcome the SHP deficits for the number of SHPs ‘above’ the current recruitment, i.e., recruiting the number of SHPs equivalent to the expected annual deficit (under the current growth rate) for a particular year. Our calculations considered the direct recruitment and retention costs for SHPs. Hence, the analysis focus was on SHP salaries, and we did not consider the broader health systems investments (see Discussion ahead). We extracted annual salary data for different SHP cadres of each state from the 2019–2020 Program Implementation Plan (PIP) and Record of Proceedings (ROP) reports [24]. State reports provided information on salaries in different formats: per capita salaries, lump-sum salary budget for a cadre, etc. Some reports did not explicitly mention all SHP cadres. Each state reports data on programmatic expenditures, including salaries, differently. E.g., some states provide the salary per person while others provide lump sum numbers for the total number of personnel belonging to a cadre hired in a financial year. Where possible, we derived the average per capita salaries by dividing the lump-sum state budgets by the total number of sanctioned SHPs (see Supplementary Table S1 for details). For states that did not have salary data or whose data could not be readily converted to per capita values, we imputed values using the median value for the cadre across states with available data. Due to the sparse data availability for specialist physicians, surgeons, paediatricians, and OBGYNs at CHCs, we grouped them as ‘doctors’ to estimate an average salary. Since the focus was on regular PHCs and CHCs, we excluded salary data for mobile medical units, obstetric intensive care units, district early intervention centres, high dependency units, etc. For all states, we extracted the salary data from requested or estimated budgets. ROP and PIP reports for 2019-20 provided heterogeneous and coarse data, which compelled us to make simplifying assumptions about the allowances, increments, etc., based on the national standards (i.e., 7th Central Pay Commission). However, in reality, such salary considerations depend on the local contexts, budgets, etc. Hence, the uniform application of national standards may not produce precise estimates; however, they do not bias the numbers in any systematic way. To understand the influence of simplifying assumptions, we conducted extensive sensitivity analyses (see salary models ahead). We calculated representative statistical salaries for SHPs. SHPs are supposed to be a grouped cadre, consisting of doctors, nurses, and ANMs. Hence, we calculated a weighted average salary for a unit of the group by multiplying the number of available SHP sub-cadres (i.e., either doctors, nurses, or ANMs) by their sub-cadre-specific salary and dividing it by the total available number of SHPs (Eq. 9). The sub-cadres considered in a ‘unit’ of the SHP groups varied across states depending on the sanctioning based on IPHS thresholds and local planning contexts. For instance, one state might need to sanction more doctors and nurses to achieve the IPHS threshold and have enough ANMs. While another state might need nurses and ANMs, it has adequate doctors. Hence, the unit costs for two such states would be different. Such differences are data-dependent and not driven by differences in definitions of the subcadres considered as SHPs. The national and state/UTs scale-up cost estimates were calculated by summing salary costs per state for each year until the given state had a deficit that year.9\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\textrm{SHP statistical salary} = \frac{{\sum}_{p}^{}({N}_{p}\times{\textrm{Salary}\:}_{p})}{{\sum}_{p}^{}{N}_{p}}$$\end{document} Where, p = sub-cadre of SHP (doctor, nurse, or ANM). N = available number of SHPs in a cadre. Salary was composed of basic pay and an allowance (28% of basic pay as per the 7th Central Pay Commission) (Eqs. 10–12) [25, 26]. 10\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\textrm{salary} = \textrm{basic pay} + \textrm{allowance}$$\end{document}11\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\textrm{allowance} = 28\% \times \textrm{basic pay} = 0.28 \times \textrm{basic pay}$$\end{document}12\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\textrm{increment} = x \% \times \textrm{allowance} = y \times \textrm{allowance}$$\end{document} Where, x = 0, 1, 5 or 10. y = 0, 0.01, 0.05 or 0.1. We used four models of salaries. These salary models were the simplest zero-order and first-order approximations of the salary structure in India. The ‘basic’ model had the annual recruitment and retention salaries (i.e., salaries required to keep the newly recruited SHP by 2030) constant over the years (Eqs. 10–11). In the increment model, the retaining salaries had an additional 1%-10% annual increment in allowance (Eq. 12) [26]. The retention period was the period from the year when recruited to 2030, as all thresholds are targeted to be achieved by 2030. We first calculated the yearly deficit for each year by taking the difference between the projected deficits for the given year and its preceding year. Next, we calculated the costs for recruiting and retaining this batch of SHPs to overcome the yearly deficit, until 2030, for all the models (Eq. 13). Finally, we summed the costs for SHP batches from 2019 to 2030 (Eq. 14). These costs exclude and are independent of the regular SHP recruitment as per current AAPC rates, i.e., these are additional costs.13\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\eqalign{ {\rm{SHP}}\,{\mkern 1mu} {\rm{batch}}{\mkern 1mu} \,{\rm{cos}}{{\rm{t}}_{\it{n}}}{\rm{ = }} & \cr {\rm{\Delta }}{\mkern 1mu} \,{\rm{defici}}{{\rm{t}}_{{\it{n - (n - \rm1)}}}} & \cr {\rm{ \times }}{\mkern 1mu} \,{\mkern 1mu} {\rm{[(basic}}\,{\mkern 1mu} {\rm{pay}}{\mkern 1mu} \,{\rm{ \times }}{\mkern 1mu} \,{\rm{(1 + }}{{\it{n}}^{\it{r}}}{\rm{))}} & \cr {\rm{ + }}\,{\mkern 1mu} {\rm{(allowance}}{\mkern 1mu} {\mkern 1mu} \,{\rm{ \times }}{\mkern 1mu} \,\sum\limits_{{\it{m = \rm0}}}^{{{\it{n}}^{\it{r}}}} {{{{\rm{(1 + \it{y})}}}^{\it{m}}}{\rm{)]}}} & \cr} $$\end{document}14\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\textrm{Total Cost} = \sum\limits_{n}^{}{\textrm{batch SHP cost}\:}_{n}$$\end{document} Where, n = year. nr = retention period. y = 0, 0.01, 0.05, or 0.1. We estimated total costs based on deficits for three different thresholds with two different salary models for two different scenarios (aspirational and realistic). National costs were the summation of the costs for all the states/UTs facing SHP deficits over the years. We calculated costs in 2020 INR and USD (exchange rate: 1 USD = 74.102 INR) [27]. We also expressed the costs for relieving the SHP deficits in rural PHCs and CHCs as percentages of total and government health spending (THS and GHS) and gross domestic product (GDP). We obtained annual projected estimates for Indian THS, GHS, and GDP from 2019 to 2030 from the Institute for Health Metrics and Evaluation [28]. Results Below, we present the findings for the realistic scenario in detail with brief remarks about the aspirational scenario. The Supplement provides all the figures and tables for the aspirational scenario findings. Current rural SHP density and growth rates We calculated the rural SHP densities for India and its 35 states/UTs from 2009 to 2030. Figure 1 depicts the changes in SHP densities from 2009 to 2019 and 2030 under the realistic scenario (i.e., SUP-adjusted population). Nationally, SHP density increased from 7.55 in 2009 to 8.85 in 2019. In 2019, among states/UTs, Mizoram had the highest SHP density of 22.52, while Odisha had the lowest density (5.22). We observed similar SHP density trends for the aspirational scenario (Supplementary Figure S1). Fig. 1SUP-adjusted SHP densities under the realistic scenario. Legend: The 2009 (past), 2019 (current), and 2030 (projected) densities (per 10,000 people) are represented by gray, blue, and green circles, respectively. States/UTs with * and ** indicate the use of 2010 and 2011 data, respectively, instead of 2009 (grey circles) SUP-adjusted SHP densities under the realistic scenario. Legend: The 2009 (past), 2019 (current), and 2030 (projected) densities (per 10,000 people) are represented by gray, blue, and green circles, respectively. States/UTs with * and ** indicate the use of 2010 and 2011 data, respectively, instead of 2009 (grey circles) Nationally, the rural SHP numbers grew at an APC of 4.51% from 2009 to 2013, shrank by an APC of 1.48% from 2013 to 2016, and then rose again at an APC of 4.22% from 2016 to 2019. Overall, the AAPC from 2009 to 2019 was 2.59% (95% CI: 0.93–4.28) (Fig. 2). The AAPC values varied across states/UTs, with 26 states/UTs showing positive growth rates. Jammu and Kashmir (8.54%) had the maximal growth, and Delhi (-7.51%) had the greatest negative AAPC (Fig. 3). Fig. 2Historical and projected number of SHPs for India. Legend: Historical SHP data (O) with 3 sub-trends (annual percentage changes or APCs) represented by blue, orange, and green lines. The black line represents the projected SHP estimates (X) from 2019 to 2030. The shaded region represents the lower and upper 95% confidence intervals for projected values. ‘w’ is the weight or period of trend, and ‘β’ is the beta coefficient of the trend fit. * shows significance p-value < 0.05 Historical and projected number of SHPs for India. Legend: Historical SHP data (O) with 3 sub-trends (annual percentage changes or APCs) represented by blue, orange, and green lines. The black line represents the projected SHP estimates (X) from 2019 to 2030. The shaded region represents the lower and upper 95% confidence intervals for projected values. ‘w’ is the weight or period of trend, and ‘β’ is the beta coefficient of the trend fit. * shows significance p-value < 0.05 Fig. 3Average annual percentage change (AAPC) values for the number of SHPs from 2009 to 2019. Legend: The bar represents the AAPC value, the line represents the 95% confidence interval Average annual percentage change (AAPC) values for the number of SHPs from 2009 to 2019. Legend: The bar represents the AAPC value, the line represents the 95% confidence interval Projected rural SHP density for 2030 With a growth rate of 2.59%, rural SHPs were projected to increase from 364,271 in 2019 to 482,643 in 2030, nationally. The rural SHP density will grow to 11.46 per 10,000 people in 2030. The UT of Daman and Diu (D & D) was projected to have the highest rural SHP density of 136.12, while Himachal Pradesh would have the lowest density of 5.39 (Fig. 1). We also noted similar SHP density trends for states/ UTs in Supplementary Fig. S1. Validation of SHP projections As part of the validation, we compared the number of SHPs in 2020 and 2021 derived from the RHS with our projected values for SHPs. We found Lin’s CCC values to be 0.92 [0.84, 0.96] and 0.90 [0.80, 0.95] for 2020 and 2021 (Fig. 4). Hence, the projections were in moderate agreement with the reported values. Fig. 4Validation of state-level SHP projections compared to actual SHP data for 2020 and 2021 from the Rural Health Statistics reports using Lin’s concordance correlation coefficient (CCC). Legend: The black line is the reference line depicting perfect agreement, while the blue and green lines depict correlations for 2020 and 2021, respectively Validation of state-level SHP projections compared to actual SHP data for 2020 and 2021 from the Rural Health Statistics reports using Lin’s concordance correlation coefficient (CCC). Legend: The black line is the reference line depicting perfect agreement, while the blue and green lines depict correlations for 2020 and 2021, respectively For several states, the projected SHP values in 2020 and 2021 exceeded the actual values. Even so, the direction and magnitude of the difference were often inconsistent for states across years. Additionally, we did not find clear evidence that such differences were associated with the population size or the geographic scope of the state (Fig. 4). Rural SHP deficits and required growth rates to achieve thresholds Under the realistic scenario, India needed an additional 560,481 SHPs for Acute, 1,440,618 SHPs for WHO-SDG, and 1,866,491 SHPs for India-SDG threshold to meet the different thresholds in 2019. At current rates, these deficits would reach 477,135 (Acute), 1,390,609 (WHO-SDG), and 1,832,613 (India-SDG) in 2030 (Fig. 5). We saw similar SHP deficit trends, only greater in magnitude, for the aspirational scenario (Supplementary Figure S2a). Fig. 5SUP-adjusted SHP deficits under the realistic scenario for India. Legend: Historical (O) and projected (X) deficits relative to the WHO-SDG (blue), India-SDG (red), and Acute (green) thresholds. The shaded region represents the lower and upper 95% confidence intervals for projections SUP-adjusted SHP deficits under the realistic scenario for India. Legend: Historical (O) and projected (X) deficits relative to the WHO-SDG (blue), India-SDG (red), and Acute (green) thresholds. The shaded region represents the lower and upper 95% confidence intervals for projections Among states/UTs, Bihar would have the greatest rural SHP deficit for all three thresholds in 2030 (Fig. 6). By 2030, four states were projected to achieve the Acute threshold, while one state (Kerala) was projected to cross the WHO-SDG and India-SDG thresholds. Two UTs (Andaman and Nicobar Islands and D & D) were projected to cross the Acute threshold by 2030 (Fig. 6). UTs like Lakshadweep, Chandigarh, and Delhi were projected to have no rural population by 2030 and hence were not considered. Under the aspiration scenario, Uttar Pradesh (UP) and Bihar were projected to have maximum rural SHP deficits for all the thresholds in 2030 (Supplementary Figure S2b). Fig. 6SUP-adjusted SHP deficits under the realistic scenario for 2030. Legend: WHO-SDG is represented in blue, India-SDG in red, and Acute threshold in green SUP-adjusted SHP deficits under the realistic scenario for 2030. Legend: WHO-SDG is represented in blue, India-SDG in red, and Acute threshold in green Under the realistic scenario, India would need to increase its rural SHPs at an AAPC of 8.84% to achieve the Acute threshold, 15.66% for the WHO-SDG threshold, and 17.91% for the India-SDG threshold (Fig. 7). Put another way, at current growth rates, India would achieve the Acute threshold by 2056, the WHO-SDG by 2082, and the India-SDG by 2090. Under the aspirational scenario, India would need an AAPC of 16.8% for Acute, 24.12% for WHO-SDG, and 26.53% for India-SDG thresholds (Supplementary Figure S3). Or, at the current rates under the aspirational scenario, India would meet the Acute threshold by 2086, the WHO-SDG threshold by 2112, and the India-SDG threshold by 2121. Fig. 7SUP-adjusted SHP densities (per 10,000 people) under the realistic scenario with AAPCs required to achieve target density thresholds. Legend: WHO-SDG is represented in blue, India-SDG in red, and Acute threshold in green. The black line notes historic (O) and projected (X) SHP densities with the AAPC based on 2009–2019 values SUP-adjusted SHP densities (per 10,000 people) under the realistic scenario with AAPCs required to achieve target density thresholds. Legend: WHO-SDG is represented in blue, India-SDG in red, and Acute threshold in green. The black line notes historic (O) and projected (X) SHP densities with the AAPC based on 2009–2019 values Costs of recruiting and retaining rural SHPs to account for deficits Nationally, assuming a basic salary model under the realistic scenario, the costs of recruiting and retaining additional SHPs in the rural public health system to cover the deficits were 1.37T INR (18.47B USD) for the Acute threshold, 3.58T INR (48.37B USD) for WHO-SDG threshold, and 4.66T INR (62.91B USD) for India-SDG threshold by 2030. Considering a 5% increment salary model, the costs were 1.46T INR (19.64B USD) for Acute, 3.81T INR (51.43B USD) for WHO-SDG, and 4.96T INR (66.89B USD) for India-SDG threshold. We depicted the cost estimates under the realistic scenario of states/UTs in Table 1, and findings considering 1% and 10% incremental salary models in Supplementary Table S2. Table 1Cost estimates for covering SHP deficits under the realistic scenario (SUP-adjusted) based on two salary models (basic and 5% increment) per three thresholdsStateBasic modelIncrement model 5%AcuteWHO SDGIndia SDGAcuteWHO SDGIndia SDGIndia1.37T INR (18.47B USD)3.58T INR (48.37B USD)4.66T INR (62.91B USD)1.46T INR (19.64B USD)3.81T INR (51.43B USD)4.96T INR (66.89B USD)Andaman and Nicobar Islands- (-)1.44B INR (19.45 M USD)2.3B INR (31.0 M USD)- (-)1.53B INR (20.69 M USD)2.44B INR (32.98 M USD)Andhra Pradesh6.35B INR (85.75 M USD)83.35B INR (1.12B USD)120.61B INR (1.63B USD)6.76B INR (91.22 M USD)88.67B INR (1.2B USD)128.3B INR (1.73B USD)Arunachal Pradesh2.52B INR (34.05 M USD)10.14B INR (136.77 M USD)13.82B INR (186.48 M USD)2.68B INR (36.22 M USD)10.78B INR (145.5 M USD)14.7B INR (198.37 M USD)Assam88.61B INR (1.2B USD)221.48B INR (2.99B USD)286.4B INR (3.86B USD)94.26B INR (1.27B USD)235.52B INR (3.18B USD)304.51B INR (4.11B USD)Bihar164.77B INR (2.22B USD)403.5B INR (5.45B USD)0.52T INR (7.0B USD)175.2B INR (2.36B USD)428.72B INR (5.79B USD)0.55T INR (7.44B USD)Chandigarh- (-)- (-)- (-)- (-)- (-)- (-)Chhattisgarh63.8B INR (0.86B USD)160.32B INR (2.16B USD)207.37B INR (2.8B USD)67.86B INR (0.92B USD)170.55B INR (2.3B USD)220.59B INR (2.98B USD)Dadra and Nagar Haveli0.72B INR (9.66 M USD)2.07B INR (27.91 M USD)2.72B INR (36.74 M USD)0.76B INR (10.27 M USD)2.2B INR (29.69 M USD)2.9B INR (39.08 M USD)Daman and Diu- (-)- (-)- (-)- (-)- (-)- (-)Delhi1.45B INR (19.51 M USD)2.95B INR (39.86 M USD)3.68B INR (49.71 M USD)1.54B INR (20.75 M USD)3.14B INR (42.41 M USD)3.92B INR (52.88 M USD)Goa1.43B INR (19.34 M USD)4.51B INR (60.88 M USD)6.0B INR (80.98 M USD)1.52B INR (20.57 M USD)4.8B INR (64.76 M USD)6.38B INR (86.15 M USD)Gujarat50.4B INR (0.68B USD)151.21B INR (2.04B USD)199.99B INR (2.7B USD)53.62B INR (0.72B USD)160.86B INR (2.17B USD)212.75B INR (2.87B USD)Haryana21.25B INR (286.79 M USD)62.88B INR (0.85B USD)83.03B INR (1.12B USD)22.59B INR (304.91 M USD)66.87B INR (0.9B USD)88.3B INR (1.19B USD)Himachal Pradesh24.34B INR (328.48 M USD)54.95B INR (0.74B USD)69.77B INR (0.94B USD)25.87B INR (349.15 M USD)58.42B INR (0.79B USD)74.17B INR (1.0B USD)Jammu and Kashmir43.54B INR (0.59B USD)115.7B INR (1.56B USD)150.62B INR (2.03B USD)46.32B INR (0.63B USD)123.08B INR (1.66B USD)160.23B INR (2.16B USD)Jharkhand16.69B INR (225.24 M USD)38.78B INR (0.52B USD)49.47B INR (0.67B USD)17.72B INR (239.09 M USD)41.19B INR (0.56B USD)52.54B INR (0.71B USD)Karnataka48.18B INR (0.65B USD)123.33B INR (1.66B USD)159.69B INR (2.15B USD)51.05B INR (0.69B USD)131.01B INR (1.77B USD)169.7B INR (2.29B USD)Kerala- (-)21.11B INR (284.86 M USD)35.24B INR (475.5 M USD)- (-)22.46B INR (303.03 M USD)37.48B INR (0.51B USD)Lakshadweep- (-)- (-)- (-)- (-)- (-)- (-)Madhya Pradesh122.46B INR (1.65B USD)278.43B INR (3.76B USD)353.9B INR (4.78B USD)130.11B INR (1.76B USD)295.82B INR (3.99B USD)376.0B INR (5.07B USD)Maharashtra30.72B INR (414.53 M USD)99.25B INR (1.34B USD)132.41B INR (1.79B USD)32.42B INR (437.45 M USD)105.3B INR (1.42B USD)140.56B INR (1.9B USD)Manipur6.84B INR (92.27 M USD)18.35B INR (247.63 M USD)23.92B INR (322.8 M USD)7.27B INR (98.1 M USD)19.51B INR (263.26 M USD)25.43B INR (343.18 M USD)Meghalaya9.87B INR (133.26 M USD)26.04B INR (351.45 M USD)33.87B INR (457.02 M USD)10.5B INR (141.75 M USD)27.7B INR (373.86 M USD)36.03B INR (486.17 M USD)Mizoram460.91 M INR (6.22 M USD)3.95B INR (53.3 M USD)5.64B INR (76.15 M USD)490.3 M INR (6.62 M USD)4.2B INR (56.69 M USD)6.0B INR (80.99 M USD)Nagaland167.21 M INR (2.26 M USD)0.63B INR (8.53 M USD)0.86B INR (11.57 M USD)177.87 M INR (2.4 M USD)0.67B INR (9.08 M USD)0.91B INR (12.31 M USD)Odisha119.3B INR (1.61B USD)267.47B INR (3.61B USD)339.16B INR (4.58B USD)126.9B INR (1.71B USD)284.52B INR (3.84B USD)360.79B INR (4.87B USD)Puducherry1.19B INR (16.0 M USD)3.2B INR (43.21 M USD)4.18B INR (56.38 M USD)1.26B INR (16.99 M USD)3.4B INR (45.87 M USD)4.43B INR (59.84 M USD)Punjab12.21B INR (164.84 M USD)43.36B INR (0.59B USD)58.45B INR (0.79B USD)12.99B INR (175.35 M USD)46.12B INR (0.62B USD)62.18B INR (0.84B USD)Rajasthan142.55B INR (1.92B USD)377.68B INR (5.1B USD)491.46B INR (6.63B USD)151.5B INR (2.04B USD)401.37B INR (5.42B USD)0.52T INR (7.05B USD)Sikkim0.93B INR (12.6 M USD)2.81B INR (37.87 M USD)3.71B INR (50.1 M USD)0.99B INR (13.4 M USD)2.99B INR (40.29 M USD)3.95B INR (53.29 M USD)Tamil Nadu46.97B INR (0.63B USD)127.73B INR (1.72B USD)166.8B INR (2.25B USD)49.97B INR (0.67B USD)135.87B INR (1.83B USD)177.44B INR (2.39B USD)Tripura1.04B INR (13.99 M USD)2.43B INR (32.81 M USD)3.11B INR (41.91 M USD)1.1B INR (14.89 M USD)2.59B INR (34.9 M USD)3.3B INR (44.58 M USD)Uttar Pradesh158.29B INR (2.14B USD)457.08B INR (6.17B USD)0.6T INR (8.12B USD)168.38B INR (2.27B USD)486.23B INR (6.56B USD)0.64T INR (8.64B USD)Uttarakhand14.58B INR (196.77 M USD)36.09B INR (487.07 M USD)46.5B INR (0.63B USD)15.5B INR (209.14 M USD)38.37B INR (0.52B USD)49.44B INR (0.67B USD)West Bengal167.14B INR (2.26B USD)382.22B INR (5.16B USD)486.3B INR (6.56B USD)177.79B INR (2.4B USD)406.6B INR (5.49B USD)0.52T INR (6.98B USD)Footnote: USD = 74.102 INR (2020 annual estimate) Table 1Cost estimates for covering SHP deficits under the realistic scenario (SUP-adjusted) based on two salary models (basic and 5% increment) per three thresholdsStateBasic modelIncrement model 5%AcuteWHO SDGIndia SDGAcuteWHO SDGIndia SDGIndia1.37T INR (18.47B USD)3.58T INR (48.37B USD)4.66T INR (62.91B USD)1.46T INR (19.64B USD)3.81T INR (51.43B USD)4.96T INR (66.89B USD)Andaman and Nicobar Islands- (-)1.44B INR (19.45 M USD)2.3B INR (31.0 M USD)- (-)1.53B INR (20.69 M USD)2.44B INR (32.98 M USD)Andhra Pradesh6.35B INR (85.75 M USD)83.35B INR (1.12B USD)120.61B INR (1.63B USD)6.76B INR (91.22 M USD)88.67B INR (1.2B USD)128.3B INR (1.73B USD)Arunachal Pradesh2.52B INR (34.05 M USD)10.14B INR (136.77 M USD)13.82B INR (186.48 M USD)2.68B INR (36.22 M USD)10.78B INR (145.5 M USD)14.7B INR (198.37 M USD)Assam88.61B INR (1.2B USD)221.48B INR (2.99B USD)286.4B INR (3.86B USD)94.26B INR (1.27B USD)235.52B INR (3.18B USD)304.51B INR (4.11B USD)Bihar164.77B INR (2.22B USD)403.5B INR (5.45B USD)0.52T INR (7.0B USD)175.2B INR (2.36B USD)428.72B INR (5.79B USD)0.55T INR (7.44B USD)Chandigarh- (-)- (-)- (-)- (-)- (-)- (-)Chhattisgarh63.8B INR (0.86B USD)160.32B INR (2.16B USD)207.37B INR (2.8B USD)67.86B INR (0.92B USD)170.55B INR (2.3B USD)220.59B INR (2.98B USD)Dadra and Nagar Haveli0.72B INR (9.66 M USD)2.07B INR (27.91 M USD)2.72B INR (36.74 M USD)0.76B INR (10.27 M USD)2.2B INR (29.69 M USD)2.9B INR (39.08 M USD)Daman and Diu- (-)- (-)- (-)- (-)- (-)- (-)Delhi1.45B INR (19.51 M USD)2.95B INR (39.86 M USD)3.68B INR (49.71 M USD)1.54B INR (20.75 M USD)3.14B INR (42.41 M USD)3.92B INR (52.88 M USD)Goa1.43B INR (19.34 M USD)4.51B INR (60.88 M USD)6.0B INR (80.98 M USD)1.52B INR (20.57 M USD)4.8B INR (64.76 M USD)6.38B INR (86.15 M USD)Gujarat50.4B INR (0.68B USD)151.21B INR (2.04B USD)199.99B INR (2.7B USD)53.62B INR (0.72B USD)160.86B INR (2.17B USD)212.75B INR (2.87B USD)Haryana21.25B INR (286.79 M USD)62.88B INR (0.85B USD)83.03B INR (1.12B USD)22.59B INR (304.91 M USD)66.87B INR (0.9B USD)88.3B INR (1.19B USD)Himachal Pradesh24.34B INR (328.48 M USD)54.95B INR (0.74B USD)69.77B INR (0.94B USD)25.87B INR (349.15 M USD)58.42B INR (0.79B USD)74.17B INR (1.0B USD)Jammu and Kashmir43.54B INR (0.59B USD)115.7B INR (1.56B USD)150.62B INR (2.03B USD)46.32B INR (0.63B USD)123.08B INR (1.66B USD)160.23B INR (2.16B USD)Jharkhand16.69B INR (225.24 M USD)38.78B INR (0.52B USD)49.47B INR (0.67B USD)17.72B INR (239.09 M USD)41.19B INR (0.56B USD)52.54B INR (0.71B USD)Karnataka48.18B INR (0.65B USD)123.33B INR (1.66B USD)159.69B INR (2.15B USD)51.05B INR (0.69B USD)131.01B INR (1.77B USD)169.7B INR (2.29B USD)Kerala- (-)21.11B INR (284.86 M USD)35.24B INR (475.5 M USD)- (-)22.46B INR (303.03 M USD)37.48B INR (0.51B USD)Lakshadweep- (-)- (-)- (-)- (-)- (-)- (-)Madhya Pradesh122.46B INR (1.65B USD)278.43B INR (3.76B USD)353.9B INR (4.78B USD)130.11B INR (1.76B USD)295.82B INR (3.99B USD)376.0B INR (5.07B USD)Maharashtra30.72B INR (414.53 M USD)99.25B INR (1.34B USD)132.41B INR (1.79B USD)32.42B INR (437.45 M USD)105.3B INR (1.42B USD)140.56B INR (1.9B USD)Manipur6.84B INR (92.27 M USD)18.35B INR (247.63 M USD)23.92B INR (322.8 M USD)7.27B INR (98.1 M USD)19.51B INR (263.26 M USD)25.43B INR (343.18 M USD)Meghalaya9.87B INR (133.26 M USD)26.04B INR (351.45 M USD)33.87B INR (457.02 M USD)10.5B INR (141.75 M USD)27.7B INR (373.86 M USD)36.03B INR (486.17 M USD)Mizoram460.91 M INR (6.22 M USD)3.95B INR (53.3 M USD)5.64B INR (76.15 M USD)490.3 M INR (6.62 M USD)4.2B INR (56.69 M USD)6.0B INR (80.99 M USD)Nagaland167.21 M INR (2.26 M USD)0.63B INR (8.53 M USD)0.86B INR (11.57 M USD)177.87 M INR (2.4 M USD)0.67B INR (9.08 M USD)0.91B INR (12.31 M USD)Odisha119.3B INR (1.61B USD)267.47B INR (3.61B USD)339.16B INR (4.58B USD)126.9B INR (1.71B USD)284.52B INR (3.84B USD)360.79B INR (4.87B USD)Puducherry1.19B INR (16.0 M USD)3.2B INR (43.21 M USD)4.18B INR (56.38 M USD)1.26B INR (16.99 M USD)3.4B INR (45.87 M USD)4.43B INR (59.84 M USD)Punjab12.21B INR (164.84 M USD)43.36B INR (0.59B USD)58.45B INR (0.79B USD)12.99B INR (175.35 M USD)46.12B INR (0.62B USD)62.18B INR (0.84B USD)Rajasthan142.55B INR (1.92B USD)377.68B INR (5.1B USD)491.46B INR (6.63B USD)151.5B INR (2.04B USD)401.37B INR (5.42B USD)0.52T INR (7.05B USD)Sikkim0.93B INR (12.6 M USD)2.81B INR (37.87 M USD)3.71B INR (50.1 M USD)0.99B INR (13.4 M USD)2.99B INR (40.29 M USD)3.95B INR (53.29 M USD)Tamil Nadu46.97B INR (0.63B USD)127.73B INR (1.72B USD)166.8B INR (2.25B USD)49.97B INR (0.67B USD)135.87B INR (1.83B USD)177.44B INR (2.39B USD)Tripura1.04B INR (13.99 M USD)2.43B INR (32.81 M USD)3.11B INR (41.91 M USD)1.1B INR (14.89 M USD)2.59B INR (34.9 M USD)3.3B INR (44.58 M USD)Uttar Pradesh158.29B INR (2.14B USD)457.08B INR (6.17B USD)0.6T INR (8.12B USD)168.38B INR (2.27B USD)486.23B INR (6.56B USD)0.64T INR (8.64B USD)Uttarakhand14.58B INR (196.77 M USD)36.09B INR (487.07 M USD)46.5B INR (0.63B USD)15.5B INR (209.14 M USD)38.37B INR (0.52B USD)49.44B INR (0.67B USD)West Bengal167.14B INR (2.26B USD)382.22B INR (5.16B USD)486.3B INR (6.56B USD)177.79B INR (2.4B USD)406.6B INR (5.49B USD)0.52T INR (6.98B USD)Footnote: USD = 74.102 INR (2020 annual estimate) Cost estimates for covering SHP deficits under the realistic scenario (SUP-adjusted) based on two salary models (basic and 5% increment) per three thresholds Footnote: USD = 74.102 INR (2020 annual estimate) Costs were greater considering the aspirational scenario. Per the basic salary model, we found the costs of recruiting and retaining SHPs to cover deficits by 2030 were 4.34T INR (58.54B USD) for Acute, 9.43T INR (127.3B USD) for WHO-SDG, and 11.9T INR (160.64B USD) for India-SDG. Similarly, considering a 5% increment salary model, the costs were 4.61T INR (62.24B USD) for the Acute threshold, 10.03T INR (135.34B USD) for the WHO-SDG threshold, and 12.66T INR (170.79B USD) for the India-SDG threshold. We depicted the cost estimates under the aspirational scenario for states/UTs in Supplementary Table S3, and findings for 1% and 10% incremental salary models in Supplementary Table S4. For the realistic scenario, assuming basic and a 5% incremental salary model, the costs for recruitment and retaining SHPs were < 4.5% of government health spending (GHS), < 1.30% of total health spending (THS), and < 0.05% of gross domestic product (GDP) to attain the Acute threshold. To attain the WHO-SDG threshold, the costs were less than 11.7% of GHS, 3.4% of THS, and 0.12% of GDP. To achieve the India-SDG threshold, the costs were about 15.15% of GHS, 4.4% of THS, and 0.16% of GDP (Table 2). Supplementary Tables S6 & S7 note the percentage spent under the aspirational and realistic scenarios, respectively. Table 2Cost estimates for covering SHP deficits under the realistic scenario (SUP-adjusted) based on two salary models (basic and 5% increment) as proportions of different national spending valuesBudgetBasic modelIncrement model 5%AcuteWHO SDGIndia SDGAcuteWHO SDGIndia SDGGovernment Health Spending4.18%10.96%14.25%4.45%11.65%15.15%Total Health Spending1.22%3.19%4.15%1.30%3.39%4.41%Gross Domestic Product0.04%0.12%0.15%0.05%0.12%0.16% Table 2Cost estimates for covering SHP deficits under the realistic scenario (SUP-adjusted) based on two salary models (basic and 5% increment) as proportions of different national spending valuesBudgetBasic modelIncrement model 5%AcuteWHO SDGIndia SDGAcuteWHO SDGIndia SDGGovernment Health Spending4.18%10.96%14.25%4.45%11.65%15.15%Total Health Spending1.22%3.19%4.15%1.30%3.39%4.41%Gross Domestic Product0.04%0.12%0.15%0.05%0.12%0.16% Cost estimates for covering SHP deficits under the realistic scenario (SUP-adjusted) based on two salary models (basic and 5% increment) as proportions of different national spending values Discussion Summary of findings Based on the average annual growth rate of 2.59% from 2009 to 2019, India’s rural SHP density at PHCs and CHCs would grow from 8.85 in 2019 to 11.47 per 10,000 people in 2030. We expect deficits of > 450,000 SHPs for the Acute, > 1 million SHPs for the WHO-SDG, and ~ 2 million SHPs for the India-SDG threshold. According to the most conservative estimates, assuming the basic salary model under a realistic scenario, the costs of retaining and recruiting SHPs are 1.37-4.66T INR to meet different thresholds. There are notable heterogeneities in SHP densities, growth rates, and current and future deficits across states. Generally, states in northern and central India seem to have low densities and high current and future deficits. These differences could be attributed to several interrelated factors, such as large geographic and population size, weak broader social development, limited health expenditure, and inadequate focus on the NRHM. Bihar, West Bengal, UP, Madhya Pradesh, and Rajasthan need to focus on scaling up their rural SHP recruitment and retention in public health centres. While India’s overall rural population is projected to grow only at a marginal rate, nearly half of the states are expected to have a decline in their rural population, with Chandigarh and Lakshadweep projected to have no rural population, which may appear as false improvement in SHP density; therefore, past annual SHP growth rates should also be considered when planning health workforce expansion. Arunachal Pradesh and Jammu and Kashmir are among the few states that have demonstrated significant improvements in SHP density, despite an increasing rural population. Literature in context While India suffers from a health workforce shortage overall, multiple studies have noted that the challenge is exacerbated in rural areas [6, 7, 29, 30]. While data is limited for private practitioners in rural areas, the shortages have been recorded across cross-sectional assessments from different parts of the country. While multiple studies have defined shortages in terms of vacancies relative to some sort of staffing norms, investigations relative to SDG thresholds have been missing until the current study. Filling this gap is critical for at least three reasons. First, the IPHS thresholds sanction SHPs based on facility norms. For example, a PHC serving a population of 30,000 is required to have 1 GDMO, 1 CHO with a nursing background, 2 staff nurses, and 1 ANM. This equates to 5 SHPs per 30,000 population or 1.67 SHPs per 10,000 population. Similarly, a CHC serving a population of 1,20,000 is required to have 3 GDMOs, 1 Family Medicine specialist, 29 staff nurses and ANMs. This equates to 33 SHPs per 1,20,000 population or 2.75 SHPs per 10,000 population. So, if a population has one PHC and one CHC, per IPHS, they would require 4.42 SHPs per 10,000 people, which is much lower than the MDG and SDG thresholds of 22.5 and 44.5 SHPs per 10,000 population, respectively [31]. Hence, IPHS-based norms may undermine the true extent of the shortage. Secondly, the IPHS thresholds are somewhat arbitrary compared to the MDGs and SDGs in that they do not have empirically established associations with population health outcomes such as maternal and neonatal mortality rates [11]. Third, vacancy measurements, while useful for annual recruitment and other operational purposes, do not consider density relative to the underlying population or incorporate any metric of health service utilization. Hence, the status of SHP availability and its future within the rural public health system, which faces the most severe shortage and is a resort for the majority of Indians to seek free primary and secondary healthcare, has remained elusive. Our use of different globally acceptable density thresholds and focus on SHPs at rural PHCs and CHCs solved the shortcomings of past studies. There exists notable literature using a diverse suite of methods on workforce forecasting [32]. Briefly, approaches can be classified based on whether they focus on supply-side or demand-side considerations. The supply models use macroeconomic factors for workforce forecasting [33]. The demand-based models use population health and utilization metrics [34]. We calculated the average annual changes in the supply of SHPs at rural PHCs and CHCs from 2009 to 2019 for forecasting SHPs for 2030. Additionally, in the realistic scenario that considered service utilization proportions, we adjusted for the proportion of hospitalizations for all broad aliment types at government/public hospitals, including the PHCs and CHCs among rural residents [19]. Hence, our overall approach also considered demand-side factors. There is a paucity of literature on the financing needed for health workforce scale-up. Previously, a national-level analysis on the scale-up of total SHPs, not just limited to rural public health systems, also investigated the costs of opening new training institutions [35]. To achieve the WHO-SDG threshold density of 44.5 personnel per 10,000 people by 2030, the study noted a shortage of over 4 million SHPs (1.05 million doctors and 2.96 million nurses/midwives). Further, increasing production capacity to cover this shortage would cost anywhere from 3.7 to 7.1 trillion INR, under varying assumptions. However, the study also noted that for a 7.1T INR investment, there would be 5.4T INR gross value added during 2026–2030 in terms of a rise in health sector employment. Our study furthers this work to depict how much it would cost to ensure that such employment occurs predominantly within the public sector, focused on rural areas. Policy implications Rural India houses over two-thirds of the population and is served by less than a third of the country’s skilled health professionals [6]. Multiple barriers exist to recruit and retain SHPs, particularly doctors, in rural healthcare centres. These include inadequate salaries and bureaucratic delays in the release of funds, neglected infrastructure in the rural public health system, poor roadway networks that limit access to the ‘outside’ world, lack of opportunities for future education and career development, and negligible resources for supporting families, including children, among others [4, 36, 37]. To reach the rural SHP density thresholds, the Central and State Governments in India need to make recruiting and retaining SHPs in rural areas a priority. Analysis of the 1983, 2002, and 2017 National Health Policies - the apex health policy documents for India - noted a growing focus on rural-urban disparities in SHP distribution over the years [38]. The study also found no increase in rural SHP density relative to urban SHP density from 1981 to 2011, justifying the need for increased policy focus. However, such a focus will also be contingent upon the rural-urban migration and other demographic shifts. Hence, the current study agrees with the need to have a greater and more targeted policy and planning focus on SHPs at the rural public health centres. The between-state differences in SHP growth rates, densities, and deficits point to the importance of state-specific policy pathways. For smaller northern and northeastern states with relatively smaller deficits, incentivizing SHPs with an incremental salary model for retention could help. For larger states in central and northern India with greater absolute deficits, only improving salaries may not be enough to cover the deficits. A mix of policy solutions would be needed to ensure better rural SHP recruitment and retention. Several strategies that incentivize rural recruitment and retention are possible [39–41]. We discuss some strategies below that might be particularly relevant [42]. First, medical colleges, councils, and public health boards could identify new graduates interested in rural service, who could become champions for the cause. For supporting such champions, commitment to long-term investment in their career development rather than short-term bonuses is critical. Second, encouraging rural SHPs to take on leadership roles and helping them coordinate with community members, such as local public health authorities and partners from the private sector (e.g., local pharmacists) are the key to their integration. Third, the use of digital tools, such as an online portal that makes the placement of SHPs transparent, is critical for gaining their trust. Such a portal should enlist all available rural placement options, allowing applicants to choose based on their needs and preferences. Imperatively, as the role of health information systems expands, cross-sectoral efforts to improve internet connectivity and expand uptake of digital devices are essential. This would enable dynamic program performance reviews, thereby aiding the process of recruitment and retention of SHPs. Finally, for states with large deficits and limited financial resources, pragmatic interim alternatives are necessary. This could be achieved by either expanding teleconsultation services to rural areas to curb the immediate shortage of specialised care [43] or upskilling other health professional cadres, like physicians trained in traditional and indigenous medicine, technicians, pharmacists, and community health workers through short-term competence-based training curricula. For instance, recently, there have been initiatives to curb the shortage of health professionals, more broadly. Community Health Officers (CHOs) have been introduced at the sub-centre level under the Ayushman Bharat program as non-physician practitioners. CHOs are trained nurses or Ayurveda practitioners who have received additional training to provide comprehensive primary healthcare (CPHC) services. CHOs are supposed to be placed at Ayushman Arogya Mandirs, which are an umbrella of sub-Centres, PHCs, urban PHCs, and health and wellness centres providing comprehensive primary healthcare. Though CHOs were introduced as an independent cadre in 2018 with an aim of improving clinical care received by patients and enhancing community-level health promotion, CHOs face significant challenges related to job security, unclear accountability due to overlapping roles with other practitioners, and inadequate incentives [44]. Hence, having clear cadre-specific training guidelines and roles and responsibilities for each level of healthcare provider would not only enhance their skills but would also significantly help with service delivery and reducing workforce gaps. Rural SHP deficits are not unique to India. Other low- and lower-middle-income countries in Asia, Africa, and South America also face the problem, and some have come up with creative solutions [1]. For instance, Thailand shifted resources from cities to rural districts, improving accessibility of services in rural areas. The country’s health system also preferred recruiting students from rural provinces for healthcare training who were sent back to their home provinces/districts under contracts with the provincial health offices, increasing the availability of rural SHPs. Their promotion was encouraged. The strategies used improved the distribution of doctors to a certain extent, as evident from the increase in the number and proportion of doctors at the district level. However, the fragmented, uncoordinated, inconsistent, and sometimes irrational application of strategies resulted in a system that cannot attract doctors to stay in rural areas when there are strong economic incentives in the urban private sector [45]. Such contracts with fresh graduates are not uncommon across the Indian states and are typically called bonds or bonded services. Bonds are an example of ‘stick’ as they compel SHPs to serve for short terms in rural areas with heavy monetary penalties for failing to serve. Similar to the caveats mentioned for the Thailand example, bonds in India have seen limited success in relatively wealthier Indian states [46]. ‘Carrots over sticks’ would work better for retaining SHPs in rural India. While the current study focuses on the rural deployment and retention of SHPs, from a health financing perspective, there is growing concern around the rising costs of medical and nursing education in India. In part, the rise is driven by the increasing privatization of education [47]. Higher education and training costs have implications for the employment choices and preferences of newly graduated SHPs. Essentially, SHPs are inclined toward employment options that can help them recover the costs. Ideally, expanding free or subsidized education in public institutions could help resolve the issue. Regardless, policies that promote deployment and retention of SHPs in rural public health systems should consider the burden of such costs as well. Policymakers should also consider that the current study shows that even though the absolute costs to recruit and retain rural SHPs might seem large, they constitute a small increase in government health spending: less than 15% across all salary models and thresholds under the realistic scenario. Since these costs are for the ten years up to 2030, the annual costs to recruit and retain rural SHPs are even smaller. For the 2021-22 financial year, the National Health Mission had a budget of over 36T INR, of which over 12T INR were allocated to health system strengthening [48]. This allocation needs to increase further to recruit and retain additional SHPs, making the additional costs to scale up the recruitment and retention of rural SHPs feasible. Strengths and limitations The study has some notable strengths. We present novel estimates for rural SHP scale-up until 2030 based on data-driven models using decade-long trends at national and sub-national levels under various scenarios. The aspirational scenario considered that the entire rural population would seek primary and secondary care at rural PHCs and CHCs. While the realistic scenario accounted for the current public health sector utilization. Hence, we conducted a range of sensitivity analyses to ensure the robustness of our findings. Further, beyond describing the number of SHPs needed, this is also the first study to estimate the costs to retain and recruit SHPs at rural public health centres. For assessing deficits, we considered three thresholds that align with global and national targets and recommendations. Use of these thresholds, instead of the IPHS norms, could ensure that sanctioning of SHPs in India’s rural public health system is aligned with India’s efforts to achieve SDGs. Conducting this exercise at the state level was particularly important, since recruitment of SHPs and allocation of funds typically occur at the state level. Most importantly, we focused on rural areas, since our past work and that of others demonstrated major inequities in SHP distribution for rural areas. Our focus on the public sector is aligned with the principle that it is fundamentally the responsibility of national and state governments to provide equitable access to healthcare, particularly at the primary and secondary levels of care, for all Indians. Our study also has several limitations. First, our AAPC values were based on data until 2019, which could have made the projections less accurate than those relying on more recent data. However, we wanted to keep the 2020 and 2021 data aside for validation. Second, we used only single-time point data (2017-18) for SUP adjustment in the realistic scenario to account for the percentage of people utilizing the public health services [19]. Ideally, annual adjustments would have been more appropriate. However, such data were unavailable. Also, adjusting for time-varying covariates is analytically challenging and would have been difficult to incorporate in the currently used joinpoint models. Third, for some states, salaries for all SHP cadres were not available. In such cases, we imputed the average salaries of SHPs based on other states. Such an approximation can bias the scale-up cost estimates. However, there is no way to know the magnitude or direction of the bias in the absence of the data. Fourth, our SHP scale-up cost analyses focused on salaries for recruitment and retention and did not include the training and infrastructural costs. As noted above, previous studies have estimated costs for additional medical colleges for increasing training capacity [35]. Hence, we did not include those. However, additional costs would be needed for accompanying infrastructural development at rural PHCs and CHCs. Estimating these at the state or national levels would be challenging in the absence of a more granular situational analysis of facilities. Fifth, there can be variations across health centres within states or over the years based on actual expenditures. For instance, districts with surplus revenue might provide incentives to their doctors in a certain year. Our data source and modeling strategy could not account for such ground-level variations. Future studies that focus on smaller regions (e.g., districts) and partner with the district health planners could help resolve this limitation. Finally, we could not include any cadres that were either recently introduced (e.g., CHOs introduced in 2018) or whose data collection and reporting only started recently (e.g., AYUSH), since our calculations were based on the annual changes in SHP numbers over a decade (from 2009 to 2019). Additionally, AYUSH practitioners have only made up a small proportion of total doctors, even in recent times. According to the 2019-20 Rural Health Statistics report, AYUSH formed 20.7% of total doctors at rural PHCs and 22.6% at rural CHCs [49]. So, while adding AYUSH would lower the deficits, it would not change the inference given the large SHP deficits in most states/UTs. Another rationale for excluding AYUSH is that the ‘Acute’ deficit threshold, which has the definition of SHPs, was based on the association between skilled birth attendant density and improvements in infant and maternal mortality as outcomes. AYUSH practitioners are not trained skilled birth attendants [50]. Hence, it would not align well with the Acute threshold and perhaps other thresholds that subsequently came after that. Conclusion Our analysis predicts a critical shortage of SHPs ranging from 450,000 to 2 million SHPs in rural India by 2030. The cost of recruiting and retaining these SHPs would be substantial, ranging from 1.37 to 4.66 trillion INR. However, these costs are quite feasible, as they would require an increase of < 15% in government health spending. However, our findings had a key limitation of a lack of transparent and publicly available data at the grassroots, i.e., health facility/district levels. Policy makers should focus on developing an updated healthcare workforce information system instead of using a ‘one size fits all’ approach like the IPHS. This would help determine the allocation of SHPs by examining both the supply-side (e.g., number of vacancies) and demand-side metrics (e.g., healthcare utilization rates and disease burden). Investing in this data structure would lead to targeted recruitment and efficient retention policies to efficiently close the gap in India’s rural healthcare system. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1