The goal of this implementation research project is to determine the feasibility of establishing and implementing an acceptable and robust audit system with community representation at secondary health facilities to improve maternal and perinatal outcomes. The implementation phases follow the standard World Health Organization (WHO) audit system. The initial step includes identifying death cases for review and subsequently collecting the detailed information on the near miss and adverse event history. A mixed methods data analysis will include both quantitative components, such as identification of trends in rates and causes of death and geographic location, and qualitative components, such as analysis of modifiable factors. The use of both types of data will provide a robust analysis of the problems and aid the audit team iin identifying and supporting priorities for action. The three-delay's model categorize the modifiable factors as the first delay (recognition of danger sign and care-seeking decision), second delay (identification and reaching health facility) and third delays (receiving adequate care and treatment at facilities). The audit team will make recommendations in collaboration with community representatives. The findings of the audits will be shared with the health facility authorities, program managers and community representatives to support policy and practice changes. A monthly monitoring cycle will be set up within the implementing facilities to ensure effective implementation of the audit systems.

Background:

Previous Efforts at Introducing Audit Systems in Pakistan Previous efforts to introduce facility Audit Systems with the support of WHO and other government initiatives have been undertaken. The Ministry of National Health Services, Regulations and Coordination, Government of Pakistan published national protocols with detailed operational guidelines, protocols, training materials, tools on phased implementation of the Maternal & Perinatal Death Surveillance and Response (MPDSR) to enhance reporting, recording, tracking, and auditing of deaths. Implementation of the protocols at the facility level was tasked to the provinces in the light of the Eighteenth Amendment. However, there is lack of harmonization and coordination within health systems.

We hypothesize that a locally relevant and reliable audit system developed in close consult with local community representatives and providers will provide sound evidence for supporting appropriate policy regarding the implementation of standard operating procedures for implementing standardized mortality audits

The specific research questions are:

I. Will the maternal and perinatal audits improve the reporting of deaths and near misses at the health facility and DHIS? II. Will the community-facility audit interaction improve the early referral proportions and facility births and reduce the three delays? III. Will the audit systems help better identify medical and non-medical factors of deaths and near-misses and how access to and quality of care be improved through integration of feedback loops within the health facilities and communities? IV. Does the implementation of facility based maternal, perinatal and neonatal mortality audit in combination with targeted community engagement and awareness activities improve the maternal and perinatal outcomes? To test our first hypothesis, our primary objective will seek to: i) Identify and establish audit committees at health facility level with community engagement; ii) Determine the feasibility of implementing an acceptable audit system with community representation in existing secondary health facilities and serving catchment population to improve maternal and perinatal outcomes through formative research combined with continuous evaluation and scaling up approach Our Secondary Objectives are to: i) assess intervention changes observed in first delay (care-seeking decision), second delay (identification and reaching health facility) and third delays (receiving adequate care and treatment at facilities); ii) propose possible solutions and record actions taken in order to improve access to quality of care through integrated of feedback loops within health facilities and communities.

AUDIT-SYSTEM IMPLEMENTATION PHASES (six phases) The project development is based on formative research which identified focal personnel readiness, stakeholder participation and readiness of the health facilities. Afterwards, this will provide assistance in the intervention component and collaboration. The process will include consultative meetings of project core committee, program advisory committee, director general health and district health office.

Phase I: Identification of facility leadership and community leadership The project core team and district manager identify facilities and community representatives who will coordinate and liaise with other stakeholders of the districts for the implementation of the community audit interventions.

Phase II: Establishing Audit Committee The health facility in-charge will establish an audit committee utilizing existing health committees at the Tehsil Head Quarters (THQs)Quality control committees, medical inspection committees, health welfare committees and joint health inspection committees already exist in health facilities at Matiari. Audit committees led by Medical Superintendent (MS)of the hospital, with a member from administration, and three to four consultants or medical officers from various departments (gynecologist, pediatrician, ENT specialist, pathologist, head of emergency department and head nurse), district magistrate or focal person from district health offices. Community engagement and audit system for project audit committee will be established at each facility, preferably members will be at least two obstetricians, 2 pediatricians, 1 administrator, project focal person and Community representatives as community audit representative. Responsibilities and structure of audit committee pre- described. Members of audit committee make their credentials updated every month in the sheet for future mentoring and sustainability and trickle-down trainings will be carried out by audit committee members.

Phase III: Audit with Community Engagement Community representatives (1-3 focal persons) male or female will be identified as community audit representatives. They will attend monthly audit meetings with facility-based audit committee to discuss the community perspectives particularly (one and two delays) and other issues pertaining in the negative maternal and perinatal outcomes. For enhancing community representation and necessary actions taken for the solution, LHWs, LHVS, CMWS, male community mobilizer (village/otaq leader or any other active member of village), or any other influential women in the village would be engaged for audit committee decisions. The aim is to develop coordinated discussion among community representative and health care provider at facility. They will provide timely oversight, monitor and response to the adverse event and later confer in the Audit meetings.

Phase IV: Training of the Audit Committee

Training for audit representatives, members of audit committee, will be conducted. 5-days training workshop to include:

Day 1: Maternal death and near misses, perinatal, neonatal deaths and morbidity outcomes Day 2: 'Three-delay' framework and identifying modifiable factors Day 3: Mentoring on the identification of the audit committees and initiating audit implementation committee Phase V: Establishing and Launching the Audit Cycle (monthly meetings)

The WHO audit system will be adapted to assess maternal, perinatal and neonatal audits with the following steps:

Step 1. Identifying cases for review Step 2. Collecting information Step 3. Analyzing information Step 4. Recommending Solutions Step 5: Implementing changes Step 6: Monitor and Evaluation Phase VI: Quarter re-certification of Audit Committee For sustainable audit system, recognition and reinforcement is inevitable. Refresher trainings and re certification will be conducted quarterly.

DATA MANAGEMENT AND ANALYSIS The data collection tool will be developed in English and then translated into the local language to be administered in the local community visiting health facilities. Data will be collected manually and entered through a specifically designed app. The app will be installed on laptop/desktop at each health facility under the supervision of project focal person. The AKU- Data Management Unit- DMU will develop the electronic database that includes the filters and data quality check indicators. Data will be entered at the end of every week and uploaded at AKU -DMU server. All data files will be stored for 5-7 years and then deleted according to organizational procedures for the permanent destruction of electronic and paper data.

The data will be analyzed through STAT v.16 and password-protected excel files. Descriptive statistics will be used to estimate maternal, neonatal, and other key quantitative variables. Three-delay's framework will be used to categorize the identified modifiable factors. The action report will be submitted based on the recommendations proposed by Audit committee.

Inclusion Criteria:

• Married women aged 15-49 years
• Children Under five years of age

Exclusion Criteria:

• Women who are non- residents of the study district and who do not provide consent.