Do health policies address the availability, accessibility, acceptability, and quality of human resources for health?: Analysis of three decades of National Health Policy of India

Background: Human Resources for Health (HRH) are crucial to improve health services coverage and population health outcomes. The World Health Organisation (WHO) promotes four dimensions - availability, accessibility, acceptability, and quality (AAAQ) for HRH strengthening. Integrating AAAQ dimensions in policymaking is essential to reduce the critical shortage of HRH in India. Methods: We created a multilevel framework to evaluate the incorporation of AAAQ dimensions along with strategies and actions that can improve them in HRH-related policies. HRH-related recommendations of all versions of the National Health Policy of India (NHPI) were classied according to targeted dimensions and cadres. We evaluated the extent to which NHPIs incorporated AAAQ dimensions over three decades. Furthermore, dimension-wise normalized indices were formulated to calculate HRH decits for pre-NHPI years to assess situational deciencies. Finally, we evaluated whether or not the HRH recommendations of NHPIs addressed the decient cadres and dimensions for the corresponding year. Results: We observed that HRH availability and quality were focused more in NHPI compared to accessibility and acceptability. Doctors were prioritized over auxiliary nurses-midwives and health assistants. AAAQ indices showed decits in all dimensions in almost all cadres over the years. The cadres focused by NHPI recommendations did not completely correspond to the decient cadres. Conclusion: The framework and indices based method can help identify the gaps between targeted and needed dimensions and cadres for effective HRH strengthening in countries. At the global level, the application of framework and indices will allow a comparison of strengths and weaknesses of HRH-related policies of various nations.


Background
Human resources for health (HRH) are "all people primarily engaged in actions with the primary intent of enhancing health".(1) HRH is a crucial component of health systems to improve health services coverage and population health outcomes.(2) HRH strengthening is, therefore, quintessential to achieve Universal Health Coverage (UHC)(3) within the broader Sustainable Development Goals (SDGs) framework.(4) However, there is a shortage of 6.9 million and 4.2 million skilled HRH in South-East Asia and Africa, respectively. (5) To address this shortage, under SDG-3.c, the World Health Organization (WHO) encourages nations with an HRH crisis to create and implement national and local policies focused on four HRH dimensions -availability, accessibility, acceptability, and quality (AAAQ). (5) India is facing a critical shortage of HRH with only 160 skilled health workers per 100,000 people. (6) In 2016, 36% of total HRH served in rural areas which had 71% population with doctors and nurses constituting the largest portion of HRH. (7) In response to global calls and the existing crisis, India adopted the target of achieving 550 physicians, nurses, and midwives per 100,000 population by 2030 under SDG indicator 3.c. (8) The National Health Policy of India (NHPI), arguably, the most comprehensive policy securing the health of a billion Indians, encompasses recommendations and plans to attain UHC.(9) The rst NHPI (1983) was framed as a response to the 1978 Alma Ata Declaration. Subsequently, NHPI was revised in 2002 and 2017 to match the country's needs and progress. Integrating HRH recommendations aimed at AAAQ dimensions in NHPI can bridge national priority-setting and global calls for action.
Analysis of NHPIs in the context of HRH is crucial for the following reasons. First, it helps to study policy changes and their effects on HRH strengthening over an extensive period of three decades. Second, it helps to evaluate the nation's policy commitment towards addressing HRH needs and achieving development targets. Third, it could guide the development of future policies geared towards areas of need in an evidence-based fashion. Thus, our analysis has three main aims: 1. Identify the degree of incorporation of WHO's AAAQ dimensions in NHPI's HRH-related recommendations.
2. Identify dimensional de ciencies in HRH cadres prior to three NHPI adoptions to gauge situational needs and also assess longitudinal trends in de ciencies.  Framework for Policy Analysis: We created a three-tiered framework with dimensions, strategies, and actions for assessing HRH strengthening (Fig. 1). The rst level is based on four HRH-AAAQ dimensions proposed by the Global Health Workforce Alliance (GHWA)(14) -availability (total number of personnel: stock and production), accessibility (distribution across sectors: public and private, demographics: rural and urban, and levels of care: primary, secondary and tertiary), acceptability (compatibility with the population being served in terms of HRH sex/age composition, skill-mix, cultural awareness, attitudes, and behaviors), and quality (competencies, training and skills, knowledge, and professional work ethics and patients' satisfaction). Based on an iterative scoping review of literature for Indian HRH, we added strategies and actions that could improve each dimension. Strategies are the broad approaches that can be used independently or in combination to improve a dimension. Cutting across dimensions and working in synergy, actions under strategies are directly implementable measures to produce such an improvement. The framework has four dimensions, twenty strategies, and twenty-four actions (Fig. 1).

Indices for Dimensional HRH De cits
We formulated four indices that quanti ed an important strategy corresponding to each AAAQ dimension. Availability de cit (AvD) measures the de cit in the total stock of HRH, accessibility de cit (AsD) measures the maldistribution of rural HRH against their gross national presence, acceptability de cit (ApD) measures the skew in skill-mix, and quality de cit (QD) measures the de cit of quali ed HRH.
AvD measures the de cit of present HRH with respect to the requirement thresholds.
AsD measures the de cit of HRH present in rural areas with respect to the national estimates.
ApD measures the de cit of nursing and supporting cadres with respect to doctors.
QD measures the de cit of quali ed HRH with respect to total estimates i.e. quali ed and unquali ed. We included only the cadres with P values available for at least two out of three years at the national level and R prescribed by at least one committee, resulting in eight cadres -auxiliary nurse midwives (ANMs), health assistant females (HAFs), pharmacists, nurses, dentists, indigenous medical practitioners -AYUSH (Ayurveda, Yoga, Unani, Siddha, Homeopathy), and doctors (graduates and postgraduates in allopathic medicine). We considered ANMs, HAFs, and nurses together as nursing cadres, while pharmacists and nursing cadres together as supporting cadres.

Analysis 1
To evaluate the incorporation of AAAQ dimensions and focussed cadres, we screened all NHPI sections for HRH-related recommendations. Each recommendation was coded according to -addressed dimensions, targeted cadres, recommended strategies, and employed actions (Additional le 1- Table 2).
Recommendations not speci c to any cadres were coded as 'non-cadre speci c'. Cadres apart from the eight mentioned above were grouped as 'others'. For each NHPI, we recorded the total number of dimension-wise and cadre-wise recommendations independently and further cross-tabulated them. Variations in the total number of mentions across NHPIs were used to determine trends in changing focus. The focus was determined by the number of mentions.  To check whether NHPI recommendations targeted the de cient HRH dimensions and cadres, we compared focused cadres for each dimension in each NHPI with their de cit categories for corresponding years. We used (a) cadre-wise recommendations corresponding to each dimension from Analysis 1 (b) cadrewise de cit categories under each dimension taken from Analysis 2.b. We arranged the cadres in descending order according to the number of NHPIrecommendations. We then compared whether or not the focused cadres in NHPI corresponded with their de cit categories i.e. whether cadres with high de cit had more focus in the NHPI recommendations and vice versa.

Dimension-wise and cadre-wise distribution of NHPI recommendations
The total HRH-related recommendations increased from twelve in 1983, thirty in 2002 to ninety in 2017. Figure 2A shows the distribution of recommendations across AAAQ dimensions. Across the years, the dimensional focus shifted from HRH quality to availability and back to quality. Focus on accessibility increased while acceptability was consistently neglected. The commonly proposed strategies for improving availability, accessibility, and quality were: establishing new and expanding existing training institutes, task shifting/sharing, and need-based changes in training, respectively ( Table 1). The total cadrespeci c recommendations increased from ve in 1983, twenty-six in 2002 to eighty-two in 2017 (Fig. 2B) with seven, four, and eight non-cadre speci c recommendations, respectively. Doctors were prioritized across all NHPIs while ANMs and HAFs were neglected. The commonly proposed implementation strategies for doctors were mandatory rural service and changes in the curriculum. Improvement in training, integrating AYUSH in underserved areas, and establishing professional councils were common strategies for nurses, AYUSH practitioners, and paramedics, respectively (Table 1). Cross-tabulation of cadrewise and dimension-wise number of recommendations is given in Table 2.

2.a. Longitudinal trends of the dimension-wise de cits over three decades
Over three decades, national AvD showed a decreasing trend for all cadres under both norms. The steepest decrease indicated as percentage drop was seen for pharmacists (150%) and doctors (766%) using Bhore (Fig. 3A) and HLEG norms (Fig. 3B) AsD increased for remaining cadres (Fig. 4). ApD showed a decreasing trend under both requirement thresholds at national (Figs. 5A and 5B) and rural levels ( Figs. 5C and 5D). QD for all cadres declined with the fall being steepest for dentists (621%) and least for pharmacists (47%) (Fig. 6).

2.b. Retrospective situational analyses of HRH for pre-NHPI years
In 1982, all cadres had critical-to-high AvD with the greatest de cit for nurses using R-1 as norm (Fig. 3A). High and moderate ApDs were noted for nursing and supporting cadres, respectively (Fig. 5A). AsD and QD were not measured due to a lack of data. In 2001, all cadres except pharmacists had positive AvDs with nurses being the most de cient (Fig. 3A). All cadres showed poor accessibility with AsD being greatest for nurses and least for HAFs (Fig. 4). For all cadres, unquali ed HRH was more than the quali ed indicating de cit in quality with QD being highest for pharmacists and lowest for AYUSH (Fig. 6). Moderate and low ApDs were noted for HRH, considering nursing and supporting cadres, respectively (Fig. 5A). In 2016, using R-2, all cadres except ANMs and HAFs had surplus availability (Fig. 3B). All cadres showed accessibility de cits with AsD categories almost identical to that in 2001 (Fig. 4). QD was highest for pharmacists and lowest for dentists (Fig. 6). Nursing and supporting cadres showed high and moderate ApDs, respectively (Fig. 5B).
3. Comparison of cadre-wise de cit categories with cadre focused in NHPIs Table 3 shows that no cadre-speci c recommendations were made under AAAQ-dimensions in NHPI-1983 while majority cadres had moderate-to-critical de cits. In 2002, no availability-related recommendations were made for HAFs and ANMs which had high and moderate AvDs. Pharmacists and nurses with critical AsD had only one accessibility-related recommendation each, and AYUSH, dentists, and ANMs with moderate-to-high de cits had none. Pharmacists and nurses with critical QDs had only two and one quality-related recommendations respectively, and AYUSH and dentists with high to moderate de cits had none. In 2017, there were twelve availability-related recommendations for doctors who had extreme surplus while HAFs and ANMs with high to low availability de cits had no recommendations. Nurses and pharmacists with critical AsD had two and one recommendations, respectively while dentists, ANMs, and HAFs with critical to high de cits had none. Doctors had fourteen quality-related recommendations despite showing a low de cit. In all NHPIs, negligible acceptability-related recommendations were made despite consistent de cits.

Discussion
Our results showcase that NHPIs' HRH-related recommendations were not structured to incorporate AAAQ dimensions. Over three decades, the main focus has been on HRH availability and quality, with reduced attention to accessibility and acceptability. Recommendations consistently seem to prioritize doctors over other cadres like ANMs and HAFs. The situational analysis revealed dimension-wise de cits in most cadres for all three NHPIs. Longitudinally, the magnitude of de cits declined for all dimensions except accessibility. NHPIs did not always focus on the de cient cadres thus failing to fully address the situational HRH de ciencies.
The 2014 GHWA report indicated that there was 'insu cient data' for reliance on India's HRH policy on the AAAQ dimensions (14). We evaluated the HRHrelated recommendations of NHPIs for the incorporation of AAAQ dimensions to ll this gap. De ciencies in HRH dimensions have been previously identi ed for India as shortages in HRH availability (15), skewed urban-rural distribution depicting disparities in accessibility (15), poor quality indicated by a substantial proportion of unregistered/unquali ed HRH (15), and low nurse-to-physician ratios compromising workforce acceptability (14). While policies focusing on increasing HRH production have had success in enhancing availability, they have failed to improve the accessibility and quality of the workforce (16).
Previously, HRH availability and accessibility in India (7) and other countries have been measured cross-sectionally (18) and longitudinally (19,20), using density as a metric. Here, we formulated normalized cadre-wise AAAQ de cit indices for HRH that can bridge research and policymaking. The advantage of indices over density is that they are indicative of surplus or de cit with respect to contextual requirement norms (thresholds) determined by factors such as health nancing, workforce e ciency, etc. In other studies, HRH-related policies have been assessed through interview methods (20), process tracing (18), or qualitative summarization (21). Our framework-based policy analysis is systematic and allows evaluation against quantitative indices. It can be used to study the progress of the National Health Mission (India) and translated to evaluate the HRH policies of other countries. The framework focuses on standardized policy analysis removing stakeholder bias seen in interviews and the investigator's bias in qualitative summaries. The conjoint application of framework and indices generates information useful for rigorous process-tracing.
Focusing on all four dimensions is crucial because population health outcomes can improve only when the high-quality HRH is available where needed in the form acceptable to the people (22). Integrating AAAQ dimensions in HRH parts of the health policies have shown to increase HRH density and improve population health outcomes (18). Adopting and locally implementing policies that address these dimensions is crucial and demands several prerequisites -a) identi cation of de ciencies in cadres through dimension-wise indices (19,20), b) identi cation of dimension-wise and cadre-wise focus of HRH policies using a predetermined framework (21,23), c) assessing the mismatches between dimensional de ciencies of HRH and focus of policy, d) experimentation with and evaluation of strategies and actions for their relevance, effectiveness, and cost-effectiveness (16,24). Application of integrated indices-framework approach will allow cross-national comparison of strengths and weaknesses of HRH policies with respect to contextual needs and generate evidence for strategies and actions instrumental for dimension-wise HRH improvements.
In India, policies should be restructured to incorporate strategies addressing AAAQ dimensions for HRH strengthening. This will allow for an evidence-based, coordinated, and sustained response towards the HRH crisis. Indicators for HRH acceptability should be to be de ned in the Indian context and actively integrated into policies. Cadre-speci c recommendations should focus on primary-care level cadres like ANMs, HAFs, community health workers, and rural physicians. Implementation of strategies like reducing attrition in HRH, removing administrative barriers for deployment, developing socio-cultural competence, and regular assessment of in-service staff is crucial. Robust efforts to continually develop and check HRH requirement thresholds are necessary to address the relevance of AAAQ indices. There is an urgent need for a dedicated national HRH policy that measures needs, and develops, executes, and monitors plans to enhance AAAQ of HRH in India.
Our study has several limitations. First, the P values for 1982 are calculated using data from different sources (Additional File 1- Table 1) demanding caution for interpretation of the results. Second, 1982 rural AvD, rural ApD, AsD, and QD were not calculated due to a lack of cadre-wise rural HRH data for the preceding year. However, following the longitudinal declining trend of the de cit, we can infer that the above-mentioned de cits must be higher than those in 2001 and 2016 for all dimensions. Third, focused dimensions and cadres in NHPIs were determined using the number of mentions. There are other ways to determine focus such as measuring relevance, effectiveness, and cost-effectiveness of the recommended strategy (24). However, such data is not for LMICs, and measuring the strength of strategies/actions was out of the scope of study (24)].

Conclusions
The framework and indices based method can help identify the gaps between targeted and needed dimensions and cadres for effective HRH strengthening in countries. At the global level, the application of framework and indices will allow a comparison of the strengths and weaknesses of HRH-related policies and indicate implementation strategies and actions.

List Of Abbreviations
HRH: Human Resources for Health AAAQ: Availability, Accessibility, Acceptability, and Quality NHPI: National Health Policy of India Strategies and actions for HRH strengthening for availability, accessibility, acceptability, and quality dimensions. Legend -HRH: Human Resources for Health.