Development of a framework of intervention strategies for point of care quality improvement at different levels of healthcare delivery system in India: initial lessons

Background Inadequate quality of care has been identified as one of the most significant challenges to achieving universal health coverage in low-income and middle-income countries. To address this WHO-SEARO, the point of care quality improvement (POCQI) method has been developed. This paper describes developing a dynamic framework for the implementation of POCQI across India from 2015 to 2020. Methods A total of 10 intervention strategies were designed as per the needs of the local health settings. These strategies were implemented across 10 states of India, using a modification of the ‘translating research in practice’ framework. Healthcare professionals and administrators were trained in POCQI using a combination of onsite and online training methods followed by coaching and mentoring support. The implementation strategy changed to a fully digital community of practice platform during the active phase of the COVID-19 pandemic. Dashboard process, outcome indicators and crude cost of implementation were collected and analysed across the implementation sites. Results Three implementation frameworks were evolved over the study period. The combined population benefitting from these interventions was 103 million. A pool of QI teams from 131 facilities successfully undertook 165 QI projects supported by a pool of 240 mentors over the study period. A total of 21 QI resources and 6 publications in peer-reviewed journals were also developed. The average cost of implementing POCQI initiatives for a target population of one million was US$ 3219. A total of 100 online activities were conducted over 6 months by the digital community of practice. The framework has recently extended digitally across the South-East Asian region. Conclusion The development of an implementation framework for POCQI is an essential requirement for the initiative’s successful country-wide scale. The implementation plan should be flexible to the healthcare system’s needs, target population and the implementing agency’s capacity and amenable to multiple iterative changes.


INTRODUCTION
As the world moves from millennium development goals to sustainable development goals (SDGs), 1 achieving SDGs require a system thinking approach. 2 A system thinking approach is one of the weakest links in the health systems of low-income and middleincome countries (LMICs). 3 This weakness is further compounded by the challenge of the low quality of care (QoC) in these health systems, a significant bottleneck for ensuring universal health coverage.
India contributes the most significant chunk of the global neonatal, under-five and maternal mortality. 4 The number of qualified doctors and combined midwives, nurses and doctor's ratio per 10 000 population are 3.3 and 6.4, respectively, compared with 23 as advocated by the WHO. 5 Only half of all the country's neonatal units have a fair number of trained doctors and nurses deployed. 6 In the last 10 years, the country's health infrastructure and resources have seen a tremendous expansion after introducing a central government-sponsored National Rural Health Mission Programme, which is now a part of the National Health Mission (NHM). 7 However,human resource availability continues to fall short compared with the requirements of health facilities. 8 In 2015, WHO SEARO launched the regional framework for QoC 9 that paved the way for the development of the point of care quality improvement (POCQI) method. 10 One recommendation of this framework was to create systems for building QI capacity at the health facility level. A team of doctors and nurses trained in the POCQI method by Open access WHO SEARO and USAID ASSIST scaled up the capacity building for quality improvement (QI) across India solely based on voluntary participation using an informal network structure. 11 This training coincided with developing a QI initiative for birthing areas and special newborn care units across all government health facilities in India, known as LaQshya in 2017. 12 The informal QoC network was formalised in 2018. 11 The network disseminated the knowledge and skill of QI across the country in a graded manner using the POCQI method-based multiple implementation interventions. This paper describes the creation of these implementation interventions across various health system levels in India and initial observations thereof.

METHODS Setting
The implementation exercise was carried out across all health systems (macro-level, meso-level and micro-level corresponding to national/state-level, district-level and facility-level, respectively) across 10 states in India. The implementation facilities included the primary health centres, community health centres, district hospitals, medical and nursing colleges across the public and private sectors. These implementations were carried out across India from 2015 to 2020. Details of these intervention strategies are given in table 1.

Study design
This is a descriptive observational study that used an adapted Translating Research into Practice (TRIP) framework 13 to implement POCQI methods in a local context (figure 1). Implementation process QI teams were identified using different mechanisms. The initial implementation process was driven by the voluntary participation of the local champions. The later stages were completed by nominations from the state health departments and development partners. These teams comprised healthcare professionals and workers from different health system levels and ranged from frontline community health workers to super-specialist doctors, in-service nursing professionals and nurse educators. These teams underwent training in POCQI skills and were subsequently mentored by members of the network mentioned above throughout the intervention strategies' timeframe. Mentoring was done using a mix of onsite visits and online sessions. With the onset of the COVID-19 pandemic in 2020, the online medium was widely used to continue building and mentoring the national QI teams. During the active phase of the pandemic, these online sessions led to the development of an innovative digital POCQI community of practice over the latter half of 2020. 14

Participants
Various healthcare workers, medical and nursing students, community members and various governmental, nongovernmental, national and international development partners were involved in implementing these strategies.

Approvals
Approvals were obtained taken from the respective supervising authorities, which included a facility in charges, state district officials, state NHMs and ministries of health. QI team members and mentors volunteered to take part in this exercise.

Ethical approvals
All of the changes tested in various QI intervention strategies described in this paper were about improving the implementation of widely accepted and evidence-based clinical practices. As no patient was being denied benefits from any evidence-based clinical practices, institutional review board approvals were not required.

Data collection
The implementation exercise generated data from different levels of the health system. The implementing team's primary data at the microlevel (facility) was collected using facility source documents, direct observations and patient interviews. Documentation of process and outcome indicators were predecided by the QI team in active consultation with the onsite mentor and central coordinator for the implementation process. The data were collected in Microsoft Excel 2016 sheets specially designed for easy use by the facility team. The data thus collected were cleaned through random cross-checks performed by the QI team leader and the QI mentor. The central coordinating team collected the data related to different intervention strategies at higher health systems (meso and macro) at the network level. This data were collated in active consultation with facility team leads and cleaned using inputs from stakeholders and development partners.

Analyses
We analysed the data related to the key stakeholders involved, number of facilities involved, QI projects undertaken, mentors, capacity building workshops, publications and QI resources developed during the implementation process. Additionally, the average cost incurred was calculated and mapped to the population affected by the exercise's implementation. These indicators were used to analyse the implementation process (table 2).

Patient involvement
No patients were involved in this work as the study's focus was to develop intervention strategies for the POCQI initiative. Similarly, no patients were involved in developing the research questions, outcome measures, recruitment and study conduct. The results were disseminated through experience-sharing workshops to facility teams of healthcare workers, providers, funding partners and governmental agencies.

Continued
Open access

RESULTS
The network developed context-specific intervention strategies throughout its QI implementation experience. As a result, 10 intervention strategies for scale-up and spread of QI were implemented over 2015-2020. Details of these intervention strategies are available in the online supplemental file.

Key features of the intervention strategies
These models were developed to overcome contextspecific challenges based on differences in health settings. Context variations were about-types of learners (like healthcare students and in-service healthcare professionals), linkages between facility (standalone facilitybased QI team(s) and community facility-based QI team(s) linked to teaching facility QI mentors), the geographical proximity of facilities to each other, involvement of other stakeholders like government health departments,

Evolution of framework
We implemented POCQI using various strategies to develop a rapidly developing framework for QI initiatives over 5 years (2015-2020). Various contextual factors influenced the QI programme implementation. The factors were (a) health facilities implementing QI, (b) stakeholders and their linkages and (c) level of the health system, that is, the microlevels, mesolevels and macrolevels (see figure 2). Various intervention strategies involving stakeholders across all health systems were used to introduce and sustain POCQI over 2015-2020, as shown in figures 3-5). The implementation framework developed rapidly in scale and scope over 2018-20 (figures 4 and 5) with the COVID-19 pandemic, the implementation strategy metamorphosed to a fully digital avatar (digital community of practice). The figures mentioned above clearly depict that POCQI implementation in an LMIC setting is a dynamic process undergoing a rapid evolution depending on the availability of resources, demands of the health system, needs of the target population in sync with the national and state health goals.

DISCUSSION
In 2001, the Institute of Medicine released the report 'rossing the quality chasm' that called for a redesign of health systems and defined the various quality elements in a healthcare setting. 15 This need for system redesign focused on the significance of QI in bringing these changes. 16 However, implementing QI at scale has proved to be a

Open access
challenge. 17 Numerous studies have shown that QI scale-up needs long-term leadership commitment, extensive training and support, full data recording and analysis, better human resource practices and dynamism in organisational culture for accepting new ideas. [18][19][20] The factors mentioned above are a significant challenge across all healthcare delivery systems, especially in LMICs.
Partner organisations developed a simplified approach of POCQI under the leadership of WHO-SEARO to offset these challenges. The current paper describes India's innovative intervention strategies to scale up this simplified QI approach-POCQI, across the health system's various levels, from primary care centres to teaching hospitals.

Open access
It is well known that QI implementation and sustenance is inherently problematic because of multiple factors that can affect them. 21 There is a need to develop intervention strategies across various levels of the health system. These interventions will potentially address various factors that can affect the successful uptake of QI initiatives. 22 The strategies described in this paper and the ensuing frameworks developed over 5 years (2015-2020) were through a multistage, inductive process.
As shown in figures 3-5, multiple intervention strategies were field tested across India. The lessons from this exercise led us to realise the importance of interplaying multiple factors in a health system while implementing QI initiatives. These factors are of vital significance for the successful initiation and sustenance of POCQI initiatives. The authors wish to draw attention to the fact that any attempt to develop an implementation framework for QI initiatives should be open to frequent adaptations depending on the local health settings' ever-evolving needs.
This paper shows that a single strategy may not suffice to disseminate and implement QI across health systems, especially for countries with heterogeneous health systems. Therefore, implementers should have a flexible approach to intervention strategies for QI initiatives. At times, the planned strategies might need to be modified Open access or entirely abandoned for a new one to roll out a QI initiative.
A recent scoping review of quality management models similarly highlighted a lack of implementation models for undertaking QI initiatives. 23 The initial lessons from this implementation exercise will significantly contribute to this nascent knowledge and help implementation agencies and researchers accelerate QI implementation.
Challenges in the roll out of QI implementing strategies: The suggested intervention strategies highlight the need for awareness about the context of specific factors for administrators and QI teams. It guides appropriate actions that can lead to positive outcomes in a health setting. Key challenges are described below: 1. Incomplete documentation and lack of robust datakeeping mechanisms cost significant time and energy at the level of the implementing team. 2. Supply chain issues and the nonavailability of essential equipment and resources were significant challenges encountered by the implementing teams. 3. Hierarchical or organisational barriers, lack of inter and intradepartmental communication can hamper synergistic QI efforts across interlinked clinical domains.

Frequent transfers of doctors, nurses and other HCWs
involved in the QI process often derailed the improvement team's efforts. 5. Inadequate capacity building of nurses in health facilities due to the absence of a dedicated continued nursing education programme hampered the QI initiative. 6. Lack of awareness of health facility staff regarding national and state programme guidelines and their implementation plan led to piecemeal implementation. This lack of awareness often led to confusion among facility health staff and adversely impacted the QoC provided to patients in these settings. 7. Sustenance of the QI project beyond the project duration was challenging due to a lack of resources and accountability.

Limitations
A limitation of the proposed framework is a subjective description of the implementation of QI. It is a post hoc analysis of various QI initiatives across India. Thus, a formal, detailed description of individual QI projects leading to strategies generation and framework synthesis may be lacking in this narrative. However, the same has been reported in the published literature by the network. 11 24-26 A formal impact assessment has not been carried out for the exercise described in this paper; however, the implementation process has been actively monitored concerning metrics mentioned before. Community participation was deficient in our intervention strategies, partly attributable to a lack of awareness and demand for high-quality care in the community. The intervention strategies evolved as a set of successful implementation initiatives undertaken across diverse clinical sociodemographic settings stacked together and spontaneously evolved into an implementation framework for the health system levels. Critics could view this spontaneous evolution as an exercise lacking planning and evaluation. However, it could be viewed as a blessing in disguise for the implementing team, as it gave them the freedom and flexibility to adapt, adopt or abandon in action. The implementation strategies were planned as per the project's intended objectives; however, during implementation, changes were made based on the situational analysis and challenges encountered. This resulted in a modified implementation strategy. Due to the lack of uniform implementation strategy across different models, the results are not comparable. This implementation exercise has generated valuable learnings and identified key challenges and limitations, which can be used by implementing teams in similar LMIC settings to build up implementation models of QI in challenging health settings. A recent meta-analysis concluded that models and frameworks could provide public health administrators with a choice of practical information that may be used to support capacity building efforts. 27 Similarly, the benefits that accrue from the impact of QI initiatives for the larger population make them an essential tool for health administrators to ensure cost-effective healthcare for the community, as reported by a recent systematic review. 28 What this study adds to the QI implementation paradigm ► The development of intervention strategies requires a mix of intuitive abilities, a clear understanding of local health systems dynamics, strong networking capacity, good communication skills, desire and a compassionate outlook towards patient care. ► Even in challenging LMIC settings, even without a framework and an implementation plan-QI work can be started using simple tools like POCQI. ► The implementation plan should be flexible to the needs of the healthcare system, target population and implementing agency/network capacity. It should be amenable to multiple iterative changes to make it appropriate for the local health settings' needs. ► The expenditure per million population to implement POCQI at scale in an LMIC like India is far more economical than many simple surgical procedures. The cost of implementing POCQI at scale for a target population of 1 million is approximately equivalent to the cost of a coronary bypass procedure in a private sector hospital in India. 29 ► As shown in this study, the process is labour intensive and requires sustained commitment to achieving the desired shift in the quality of delivered healthcare.

CONCLUSION
Frameworks and models help describe and understand how interventions can be scaled up 30 from small, individual health facility-based projects to a broader set of guidelines for a health system. Effective scaling up of such initiatives requires the systematic use of evidence and Open access data from on-ground implementation to drive the policy and decision-making process throughout the health system-from the national level down to the community health worker level. This paper emphasises that there is no one panacea for successfully implementing QI. Each time, the implementor has to base their choice on picking the model based on local factors 31 -preparation of the site, availability of funding, development partner or government support, availability of long-term mentoring support, geographical area of implementation and social determinants of health. This fine art of balancing macrolevel, mesolevel and microlevel contexts in a setting can significantly affect seeding and spreading QI initiatives and help in a seamless implementation. during implementation of QI at Nursing Colleges of Madhya Pradesh) and Dr Gagan Gupta (who was Health Specialist with UNICEF when the QI interventions for SNCUs at Madhya Pradesh were started) for their support and guidance during implementation of QI initiatives in various health facilities. We acknowledge the support and guidance provided by Dr Archana Verma (General Manager, Quality, NHM Uttar Pradesh) for her support and guidance during the implementation of QI initiatives in district health facilities in Uttar Pradesh. We would also like to thank all healthcare providers and workers who shared their experiences of implementing QI, which helped develop QI strategies.
Contributors VD, SS, RG, NL, KS, SKP, AS, MS, SSJ, JP, AV, VS and VB were responsible for conducting onsite visits, mentoring facilities teams and collecting and collating data from the participating facilities. RG, VD, SS, RM and NL were responsible for developing the draft manuscript and analysing the participating facilities' data. All listed authors have contributed significantly in conceptualising, designing, coordination and conduction of the intervention strategies. All authors reviewed and approved the final draft of the manuscript.
Funding Some of the implementing strategies were funded by the WHO-SEAR Office, UNICEF Madhya Pradesh and UNICEF India offices. Publication of this article is made Open Access with funding, support from the UNICEF India and Nationwide Quality of Care Network.
Competing interests RM is employed with the WHO-SEAR Office. He was responsible for the release of funds for some of the implementing strategies described in the paper. VS, VB are employed by UNICEF India and Madhya Pradesh field offices respectively, they were responsible for the release of funds for some of the implementing strategies described in the paper.The funding agencies had no bearing on methodology, data collection, data analysis and results reported in this paper. The opinions expressed herein are those of the authors and do not necessarily represent the official views of the WHO or UNICEF.
Patient consent for publication Not required.
Provenance and peer review Not commissioned; externally peer reviewed.
Data availability statement Data are available upon reasonable request. All data relevant to the study are included in the article or uploaded as supplementary information. All relevant data is available upon reasonable request from the corresponding author and also available as online supplementary material.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.
Open access This is an open access article distributed in accordance with the Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits others to copy, redistribute, remix, transform and build upon this work for any purpose, provided the original work is properly cited, a link to the licence is given, and indication of whether changes were made. See: https:// creativecommons. org/ licenses/ by/ 4. 0/.