Creating responsive health systems: improving the use of feedback from service users in quality assurance and human resource management in Bangladesh

This project aims to assist the national policymakers to make Bangladesh’s health system more responsive. It will do so through: a) assessing the current system of collecting, and responding to, feedback from health service users at sub-district (Upazila) level, and b) designing a comprehensive health systems intervention to improve utilisation of user feedback in health service quality assurance and human resource management processes at Upazila level.

Responsiveness is a widely-recognised key objective of national health systems. Responsive health systems anticipate and adapt to future health needs, and harness emerging opportunities to promote universal access to effective interventions. Effective interaction and engagement between the users of health services and service providers and managers is a vital component of responsive health systems. Two issues are central to this. First, is the opportunity for health service users to provide feedback on their experiences (e.g. care they receive, staff expertise or availability of supplies). We use the term ‘feedback’ because it includes both complaints and praise i.e. positive reflections, from service users. Second, is the ability of the health system to adequately respond to, and utilise, user feedback in e.g. improving health service quality or strengthening human resource (HR) management processes.

Since 2009, the Ministry of Health and Family Welfare (MOHFW) in Bangladesh is implementing a national program to enhance service users’ voice through allowing them to text any feedback via SMS texts. Examples of feedback include poor patient-staff interaction, out-of-stock medicines or unsanitary toilets. Health facilities in all districts have display boards showing programme information and a number for SMS texts. All texts go into a national web portal (, which has users’ names, issue and dates of receipt and of solution, and is monitored by the MOHFW. Each text is to be followed up with a phone call to both the sender and local authorities. Users can also send feedback directly to health management committees in Upazilas, and through suggestion boxes in each health facility. It is unclear, however, who and how follows up on the issues received directly at Upazila level.

Our engagement with national policymakers revealed that the implementation of this programme is patchy and needs strengthening. The ministry receives around 1,000 messages per day, however the MOHFW has only two dedicated staff to follow up on each SMS. No information is available as to the type of feedback received directly by the health management committees and to what degree the issues are addressed. As a result, ensuring responsive health system in one of the world’s most densely populated countries remains a major challenge. Our discussions with MOHFW also show that policymakers are unclear how well (if at all) these systems are integrated with, and support, service quality assurance and HR supervision and performance appraisal.

Hospital in Bangladesh (below)Hospital bangladesh

Aim & Objectives
The AIM of this project is to assist the policymakers in designing a comprehensive health systems intervention to make the Bangladesh’s health system more responsive. Key QUESTIONS that inform the aim are: what are the strengths and weaknesses of the current system of collecting and responding to service user feedback at Upazila (sub-district) level; which key contextual facilitators and constraints at the macro (health system), meso (organisational) and micro (individual) levels influence the performance of this system; and which comprehensive health systems intervention can be designed to improve the current system and integrate it with quality assurance and human resource management, in order to make the Bangladesh’s health system more responsive?

Specific project OBJECTIVES are to work closely with key national and local decision-makers to:

  • Develop an in-depth understanding of the nature and contents of, and key reasons for, feedback received from health service users at Upazila level;
  • Analyse the processes of collecting and responding to service users’ feedback at Upazila level, as well as the key contextual facilitators and constraints influencing these processes;
  • Assess the approach to, and processes of, service quality assurance and human resource management, focusing specifically on the use of feedback from service users at Upazila level;
  • Using results of objectives 1-3, develop a comprehensive and context-specific health systems intervention to improve the use of feedback from service users in quality assurance and human resource management processes at Upazila level.

The achievement of project objectives is a crucial first step of a longer-term plan to implement and assess a comprehensive intervention at larger scale to improve responsiveness of Bangladesh’s health system. The close links between the ARK Foundation and the MOHFW will facilitate this scaling up.

The study will be conducted in Comilla, a typical peri-urban district about 100km south-east of the capital, Dhaka, with a population of 5.4 million. It was selected in discussions with the MOHFW, based on frequent users feedback, existence of motivated district health leadership and our previous experience working in the district. We will purposefully-select two Upazila Health Complexes (UHC) following a review of the feedback environment in the district using observation and document review. One UHC will have a ‘favourable’ feedback environment (i.e. clear signs and processes) and one UHC a ‘less favourable’ one.

Study design
This study will be a multi-disciplinary and mixed-method health systems research, using Realist Evaluation (RE). A RE approach addresses question about what works, for whom, in which circumstances, and why. In RE researchers develop, empirically validate and refine the Middle-Range Theories (MRTs). A MRT covers how the Context (at micro, meso and macro levels) influences the intervention Mechanisms (e.g. actors’ behaviour in implementing intervention) to produce the intended Outcomes (though unintended outcomes are also identified). This is known as studying C-M-O configurations, which allow comprehensive recording all intervention aspects.

The diagram below summarises the initial programme theory, which will be continuously validated and refined in the data collection and analysis. Detailed C-M-O configurations (e.g. C1+M2=O2) will be developed, and will include specific Cs, Ms and Os (identified through objectives 1-3). These, together, will inform the design of the comprehensive health systems intervention (objective 4). 


 The study target populations will be: a) users of health services at UHC, most of whom are vulnerable groups such as women or the poor, and b) service providers and managers. The intervention, to be designed in achieving the objective 4, will include detailed guidance for each target group. For users, the intervention will detail methods for enhancing their engagement with feedback systems (e.g. improving awareness of strengths of current systems, using health management committees). For providers and managers, we will have context-specific tools to better utilise user feedback in quality assurance (e.g. critical incident technique) and HR management (e.g. revised supervision format, contents of staff performance reviews).

Study methods
REs are method-neutral, meaning that combinations of methods from different disciplines are possible. We will draw upon: process evaluations, statistics, social sciences and health policy and systems research (HPSR). Different qualitative and quantitative methods will be used for data collection and analysis. These will be updated, reflecting the evolution of specific MRTs, and will include combinations of:

  •  in-depth interviews with service users (about 20 in each UHC, which in our experience is sufficient to capture key perspectives and achieve data saturation) and focus groups with communities (2-3 in each UHC), to explore their knowledge and use of feedback systems. Participants will be purposively selected based on gender, age and use of feedback systems.
  • in-depth interviews with purposefully-selected service providers and managers (about 10 in each UHC, which in our experience is sufficient to capture key perspectives and achieve data saturation) to explore their views on and experience engaging with the user feedback systems.
  • analysis of country-level secondary data on user feedback from the web portal and UHC records, to understand types of issues, location, gender and age of users who initiated issues.
  • non-participant observation of: feedback environment in the district, health management committee meetings and UHC routine quality assurance and staff management practices.
  • review of key documents e.g. feedback to users and actions taken, meeting minutes, quality assurance guidelines, staff performance appraisal and supervision records.

The specific MRTs will be continuously refined, and will provide framework for the data collection and analysis. Analysis of qualitative and quantitative data will be done using convergent mixed methods model25 i.e. involving continuous triangulation of multiple datasets. We will work closely with decision-makers, in a research-policy partnership26, to facilitate adoption of results into policy

Any project outputs will be summarised here as soon as they are available.

In the meantime, with any questions please feel free to contact any member of the project team.

We will ‘embed’ the project into policy and practice, working with different policy actors. Decision-makers at different levels (Upazila, MOHFW) will be continuously engaged during the project in a research-policy partnership, drawing on the ARK’s extensive experience working with the government. The main beneficiaries of this project are:

  1. Policy-makers and implementers at Upazila, District, Division, National, Asian and Global levels responsible for the development and implementation of policies and systems concerning user feedback, quality of care and human resource management;
  2. Bangladesh’s public and private service providers and practitioners;
  3. Local, national and international civil society organisations and user associations with interest in advocacy and social accountability of the health systems; 
  4. National and international academia interested in health systems and policy research and evaluation of complex interventions using robust methodologies;   
  5. Local, national and international media with a focus on health and social issues.

The study results will be used to achieve improvements in: policymaking, policy implementation, health systems and health outcomes. We expect our research to have impact on policy and practice in Bangladesh and internationally, due to:

  1. improvements in the user feedback system, enhancing the responsiveness of Bangladesh’s health system and further empowering the public to hold health system to account;
  2. development of expertise in how to advise policy-makers and practitioners on the development and implementation of context-specific comprehensive health systems intervention;
  3. utilisation of an innovative cross-disciplinary approach to assess the effectiveness of a complex intervention;
  4. capacity strengthening and use of an innovative research approach in one of the world’s most densely populated countries;
  5. scientific advancement of theories on how to make health systems more responsive in the context of low- income countries.

Project Partners
The Nuffield Centre for International Health and Development, LIHS, University of Leeds:

· Dr Tolib Mirzoev (Principal Investigator) –

· Dr Helen Elsey -

· Dr Bassey Ebenso -

The Advancement through Research and Knowledge (ARK) Foundation. Bangladesh:

· Dr Rumana Huque -

· Mr Zunayed Al Azdi -

· Dr Shammi Nasreen -

Funding agency: Joint MRC / ESRC / Wellcome / DFID health systems research initiative call 4

Project period:  January 2017 – June 2018

Further information
For more information please contact any member of the project team directly. This webpage was updated on 20.4.2017.