India’s primary healthcare reform: Improved service delivery, reduced mortality

On average, a 60-year-old in India can expect to live for about 19 years, four years less than their counterparts in high-income countries – with the gap being partly driven by poor access to quality healthcare services. Based on data from Rajasthan, this article shows that adding a mid-level healthcare worker to public healthcare facilities in rural areas – a reform undertaken at scale across the country – improves service provision, increases patient visits, and decreases all-age mortality rates.

In recent decades, India has made remarkable progress in improving maternal and child health outcomes. Child mortality rates have fallen substantially, gradually converging with those seen in high-income countries. By contrast, the gap in life expectancy at age 60 between India and high-income countries persists and has even increased over time.¹A major contributing factor is poor access to quality healthcare services. According to the National Sample Survey 2017-2018, 38% of deceased adults in rural India received no medical attention before their death. Even when individuals do seek care, they often face unreliable services and low-quality treatment (Das et al. 2016).

As part of the Ayushman Bharat Initiative that was launched in 2018, policymakers in India aimed to address these shortfalls and strengthen the healthcare system by converting primary healthcare facilities into Health and Wellness Centres, now known as Ayushman Arogya Mandirs. One of the key parts of the reform was to post Community Health Officers (CHOs) to sub-centres. Sub-centres constitute the lowest level of India’s primary healthcare system and were previously only staffed by an Auxiliary Nurse Midwife (ANM) and Accredited Social Health Activists (ASHAs). The new CHOs are contractual workers who are required to have a nursing degree and mandated to provide basic outpatient care services and screening for chronic diseases to the entire community. By 2024, over 1,38,000 CHOs had been posted to facilities across the country, impacting healthcare delivery for over 750 million people (Ministry of Health & Family Welfare, 2024).

While the reform’s goal was to bring comprehensive primary care closer to rural communities, its impact on health outcomes is uncertain. Given high absenteeism among healthcare staff (Chaudhury et al. 2006), simply adding personnel might not lead to better services. Moreover, low trust in public facilities and limited information could mute patient responses (Wagner et al. 2023). If the quality of care is difficult to assess and referrals are rare, the reform may also inadvertently harm patients with complex conditions if it leads them to rely solely on mid-level healthcare workers rather than seeking care from more qualified practitioners in nearby towns.

Study design and data sources

In our study (Agte and Soni 2025), we evaluate the effects of adding CHOs to sub-centres on service provision, patient visits, and health outcomes in Rajasthan. We combine administrative data with two rounds of primary survey data from 193 villages in the Udaipur district to gain a comprehensive picture of how the deployment of CHOs affected patients and healthcare providers.

Importantly, only around two-thirds of CHO vacancies were filled in Rajasthan during the first wave of implementation in March 2022.² This staggered rollout allows us to compare changes in outcomes over time between sub-centres that did and did not receive CHOs in the first wave, using a matched difference-in-differences design. The decision of which sub-centre in a district received a CHO was made by local officials who were only given information about CHOs' places of residence and sub-centre locations. To account for this assignment process, we match treatment and control group sub-centres on observable geographic characteristics.

We rely on three data sources. Our large-scale administrative dataset is based on routine data collected by ANMs and covers all villages in Rajasthan over a four-year period (April 2019-March 2024). We further collected two rounds of original survey data on public providers, private providers, and households across 193 villages: the first just prior to the deployment of CHOs, and the second 9 to 12 months after CHOs were added to the treated sub-centres. During the second round of data collection, we also conducted unannounced audit visits and patient exit surveys at sample sub-centres to obtain additional measures of healthcare access and provider performance. Finally, in collaboration with Khushi Baby, a local NGO, we obtained data on household provider choices through an ongoing healthcare census.

Findings

We find that CHOs substantially improved public service provision. Using data from our unannounced audit visits, we show that sub-centres with a CHO are 25 percentage points more likely to be open on a given day (Figure 1, left panel). The quality of healthcare services increased as well: sub-centres in the treatment group had higher checklist completion rates in hypothetical medical vignettes and improved patient-provider interactions according to patient exit surveys (Figure 1, right panel). Results from patient exit surveys also show that patients that visited treatment group sub-centres had higher levels of satisfaction and were more likely to have their blood pressure measured.

Figure 1. Effect of Community Health Officers on healthcare access and quality

Notes: (i) The figure presents the results from unannounced audit visits and patient exit surveys. (ii) The left panel shows the share of sub-centres that were open at least at some point during the day of the unannounced visit. (iii) The right panel shows the number of questions that a healthcare provider at the sub-centre asked the patient according to patient exit surveys. (iv) The whiskers correspond to 95% confidence intervals (indicating that if we repeated the study many times, the true effect would fall within this range in 95 out of 100 cases) based on ‘standard errors’ clustered at the sub-centre. (v) Sub-centre-level weights for the control group are constructed by inverse probability weighting, which gives more weight to those sub-centres that were more likely to receive a CHO.

This translated into increased utilisation of public healthcare services (Figure 2, left panel). While treatment and control group sub-centres followed very similar levels and trends before the reform, we observe a clear trend break after CHOs were posted, leading to an average increase in patient visits by 58% in the treatment group relative to the control group. Who are these additional patients? Data on patient choices from the healthcare census suggest that this effect is driven by new patients who otherwise would not have sought any care. Sub-centres with CHOs saw a significant increase in diagnoses of both acute conditions (like heart attacks and epilepsy) and chronic illnesses (like hypertension and diabetes), indicating gains in both curative and preventive care. However, there was no change in maternal and child health services, partly because they were already being provided by the existing ANM, whereas the new CHOs mostly focused on adult healthcare.

Figure 2. Effect of Community Health Officers on patient visits and mortality

Notes: (i) The figure presents average outcomes for treatment and control group sub-centres over time. (ii) Data on patient visits and deaths are obtained from administrative data. (iii) The all-age mortality rate is defined as the number of deaths in a quarter per 1,000 people in the catchment area of the sub-centre. (iv) Sub-centre-level weights for the control group are constructed by inverse probability weighting. (v) The treatment effect is obtained from matched difference-in-differences regressions that pool all pre- and post-period quarters. (vi) indicates significance at the 5% level and indicates significance at the 1% level.

Besides directly affecting public service provision, the CHO reform also had indirect effects on private providers. Using information from private provider surveys, we show that private providers raised their quality in response to the arrival of CHOs. Consistent with a decline in market power, these effects were strongest in villages where only one private provider was present at baseline.

The reform led to measurable improvements in health outcomes as well. Using state-wide administrative data, we find that villages that are served by sub-centres with a CHO experienced a 10% reduction in all-age mortality rates over two years (Figure 2, right panel). This reduction was almost entirely concentrated among the elderly (ages 56+), translating to an increase in their life expectancy by at least three months.³Survey data further show a decline in hospitalisations, again driven by older adults. These findings are robust to various sensitivity checks and align with international evidence on the mortality impacts of expanding primary care access (Bailey and Goodman-Bacon 2015, Bancalari et al. 2023, Mora-Garcia et al. 2024).

To better understand why the reform succeeded, we estimate a ‘discrete choice model’ of patient behaviour. The model accounts for how patients value different provider characteristics, like quality, distance, and price, and simulates how they respond to changes in service delivery. Using the model, we show that the CHO reform was successful because it simultaneously improved healthcare access and quality in the public sector. By contrast, only improving either access or quality in isolation would have limited effects. The reason is that, while increasing quality is necessary to improve health outcomes, patient choices only seem to respond to changes in access but not quality. By increasing access and quality at the same time, the CHOs managed to reach new patients and also support existing patients who had already visited the sub-centre before the reform. Model simulations further suggest that accounting for local market conditions during the assignment process of healthcare workers could achieve even larger improvements in health outcomes.

Conclusion

Our analysis shows that adding CHOs to sub-centres improved service provision and health outcomes. A cost-effectiveness analysis suggests that the reform would even pay for itself in the long run by not only reducing mortality but also decreasing future government spending on hospitalisations. The impacts on private healthcare providers further demonstrate the importance of accounting for indirect effects on other stakeholders when evaluating large-scale policy reforms. Taken together, the findings demonstrate that adding mid-level providers to public primary healthcare facilities can be a low-hanging fruit to strengthen local health systems. More broadly, the results contribute to recent literature that shows how staffing expansions can improve State capacity (Ganimian et al. 2024). However, further research is needed to understand the long-run effects of the CHO reform and whether the impact of CHOs changes once they become permanent staff.

All views and errors are solely the authors’, and this article does not necessarily represent the views of the Government of Rajasthan.

Notes:

Life expectancy at age 60 increased from 16.6 to 22.9 years between 1960 and 2019 in high-income countries but only increased from 14.2 to 18.6 years in India (United Nations, 2022).
While we define the ‘treatment group’ as sub-centres that received a CHO in March 2022, 15% of ‘control group’ sub-centres also received a new CHO in December 2022, which explains the positive trend in patient visits in the control group in Figure 2.
Our administrative data provide aggregate information on deaths for five age groups: infants (<1 year), children (1-4 years), adolescents (5-14 years), adults (15-55 years), and the elderly (>56 years). We find no significant changes in the mortality rates of other age groups.