IMPROVING THE QUALITY OF MATERNAL DELIVERY AND NEWBORN CARE SERVICES THROUGH STAFF REDEPLOYMENT AND CAPACITY BUILDING AT MITYANA HOSPITAL

IMPROVING THE QUALITY OF MATERNAL DELIVERY AND NEWBORN CARE SERVICES THROUGH STAFF REDEPLOYMENT AND CAPACITY BUILDING AT MITYANA HOSPITAL BY BIZIMANA ABEL MakSPH-CDC FELLOW 2012. i

IMPROVING THE QUALITY OF MATERNAL DELIVERY AND NEWBORN CARE SERVICES THROUGH STAFF REDEPLOYMENT AND CAPACITY BUILDING AT MITYANA HOSPITAL BY BIZIMANA ABEL (MSC. HSM, UMU) MakSPH-CDC FELLOW NOVEMBER, 2012 ii

TABLE OF CONTENTS DECLARATION... vi DEDICATION... vii LIST OF ACRONYMS... viii OPERATIONAL DEFINITIONS... x EXECUTIVE SUMMARY... xii SECTION 1. INTRODUCTION AND BACKGROUND... 1 1.1. Introduction... 1 1.2. Trends in maternal and neonatal morbidity and mortality... 2 SECTION 2. THE QUALITY IMPROVEMENT PROCESS... 5 2.1. Reasons for improvement... 5 2.2. Examining the current situation... 5 2.3. Problem analysis... 7 SECTION 3. PLANNING COUNTERMEASURES... 10 3.1. Intervention objective... 10 3.2. Intervention strategies... 10 3.3. Expected results... 10 3.4. Implementation framework... 11 3.5. Reviewing staff redeployment... 13 3.6. Training Midwives at Mityana hospital... 13 SECTION 4. THE QUALITY IMPROVEMENT OUTCOMES... 15 4.1. Redeployment of midwives... 15 4.2. Training process... 16 4.3. Other changes... 17 4.4. Intervention outcomes... 18 4.5. Qualitative evaluation... 21 iii

4.6. Institutionalization of CQI... 22 SECTION 5. FUTURE PLANS... 25 SECTION 6. CHALLENGES... 26 SECTION 7. DISCUSSIONS... 30 SECTION 8. CONCLUSION AND RECOMMENDATION... 33 REFERENCES... 36 Appendix I.... 43 Appendix II... 48 Appendix III... 48 iv

LIST OF FIGURES Figure 1-1: Risk factors found among mothers attending maternity ward in Mityana hospital, 2011 (Mildmay Uganda, 2012)... 4 Figure 2-1: Fishbone analysis showing factors associated with deployment of few and less skilled midwives in maternity ward... 8 Figure 4-1: At the stakeholders meeting; the fellow makes a point as the RDC listens. 16 Figure 4-2: Participants of the stakeholders Meeting... 16 Figure 4-3: Pre and post test results for 24 midwives trained at Mityana hospital.... 17 Figure 4-7: A midwife assisting a mother... 18 Figure 4-4: Midwives and trainers ready for a practical session after 4-days theory... 18 Figure 4-5: One of the groups of trainees doing 5 pre-test... 18 Figure 4-6: A midwife gives a health talk mothers at discharge after delivery by C/S. 18 Figure 4-8: Trends of completely plotted partographs (January to November 2012)... 19 Figure 4-9: Trends of maternal and neonatal sepsis, postpartum hemorrhage and Fresh Stillbirth (FSBs)... 20 Figure 4-10: Trends in delivery methods and FSBs in relation to completely filled partgraphs.... 21 Table 2-1: Ranking of selected problems associated with poor quality of maternal delivery and newborn care in Mityana hospital.... 6 Table 2-2: Distribution of midwives in wards and departments of Mityana hospital... 7 Table 3-1: Implementation framework... 12 v

DECLARATION I Abel Bizimana do hereby declare that this programmatic report entitled Improving the Quality of Maternal Delivery and Newborn Care Services Through Staff Redeployment and Capacity Building at Mityana Hospital has been prepared and submitted in fulfillment of the requirements of the MakSPH-CDC Fellowship Program and has not been submitted for any academic qualifications. Signed. Date. Abel Bizimana, Fellow Signed. Date. Dr. Yvonne Karamagi Primary Host Mentor, Mildmay Uganda Signed. Date. MS. Mary Odiit, Secondary Host Mentor, Mildmay Uganda Signed. Date. Dr. Geoffrey Kabagambe Academic Mentor vi

DEDICATION This report is dedicated to David Wilber Rwalinda and I.K. Murekezi who unblocked my academic road. I would not be at this professional level if they did not intervene at strategic points in my career path. Late Pastor Jonathan Nkiriyehe who provided a parental and moral background that enabled me to socially fit in society. Martha Nkiriyehe, Penninah N; Jane B, My children: Munyantwali, Muhoza, Mbanjingabo, Mushakamba and Mberey ingoma and my family members whom motivate me to work hard since I have to account for my absence while pursuing my academic and professional goals. Kisoro District Local Government that has provided me with capacity and opportunity to become a public health professional; through supervision, training, allocating challenging assignments hence allowing me to advance in my studies. vii

LIST OF ACRONYMS AIDS AMTSL ANC CAO CME C/S CQI ECN EmONC FHI FIGO FSB HIV ICM IMR IRB MCHIP MDGs MMR MDNBC Acquired Immune Deficiency Syndrome Active Management of Third Stage of Labour Antenatal Care Chief Administrative Officer Continuous Medical Education Caesarian Section Continuous Quality Improvement Essential Newborn Care Emergency Obstetric and Newborn Care Family Health International Federation of International Gynaecology and Obstetrics Fresh Stillbirth Human Immunodeficiency Virus International Confederation of Midwives Infant Mortality Rate Instructional Review Board Maternal and Child Health integrated Program Millennium Development Goals Maternal Mortality Rate Maternal Delivery and Newborn Care viii

MNPI MoH MTCT MUg PATH PNO PMTCT POPPHI PPH QoC SAA SBA TQM UBOS UDHS UNFPA USAID WHO Maternal and Neonatal Program Index Ministry of Health, Uganda Mother-To-Child transmission of HIV Mildmay Uganda Program for Appropriate Technology in Health Principal Nursing Officer Prevention of Mother-To-Child transmission of HIV Prevention of Postpartum Hemorrhage Initiative Postpartum Hemorrhage Quality of Care Sub-Saharan Africa Skilled Birth Attendant Total Quality Management Uganda Bureau of Statistics Uganda Demographic and Health Survey United Nations Population Fund United States Agency for International Development World Health Organization ix

OPERATIONAL DEFINITIONS 1. Active Management of Third stage of e of Labour (AMTSL): Use of oxytocics, controlled cord traction and massaging uterus (after delivery of placenta) to prevent postpartum hemorrhage. 2. Clients: A mother, her relative and a baby in the health facilities 3. Client Participation: The process of getting clients opinion on quality of maternal and newborn care services which they consume with the purpose of using the opinion to improve the services and client satisfaction. 4. Competence: Ability to do well tasks related to ensuring health of the mother and the baby. 5. Continuous Quality improvement: Deliberate and defined processes of continuously adding on on-going efforts to achieve measurable positive change on performance, efficiency and effective delivery and immediate postnatal newborn services. 6. Delivery: A process of the childbirth that begins with labour pains and ends when the baby and the placenta have been expelled from the uterus. It involves procedures of clean birth and additional assistance to the mother, if labour does not progress well. 7. Delivery and Immediate Postnatal Newborn Care Services: These are services offered to a mother and newborn during delivery and within the first hour after delivery to prevent and control of neonatal asphyxia, MTCT, postpartum hemorrhage and sepsis. 8. Health worker: A midwife or any other competent person that participates in conduct delivery and newborn care. 9. Improvements: Changing from low quality to higher quality of maternal and newborn services x

10. Immediate Postpartum Care: The care of mother and baby aiming at ensuring that the neonate breathes and is kept warm and promptly initiated to breastfeeding, while the mother is assisted to prevent blood loss due to child birth. 11. Participatory Continuous Quality Improvement: It is a quality improvement process that deliberately and systematically engages all stakeholders involved in demand and utilization of maternal and newborn services. 12. Performance: Higher productivity and better quality of maternal and newborn services through adherence to standards of delivery and newborn care. 13. Productivity improvement: Increase in outputs in delivery of maternal and newborn care in a given time based on commitment and competence of service provider aided by existing technology. 14. Quality: The process of not only meeting clients needs (expert opinion) and expectations (clients wants) but also exceed them to attain unprecedented levels of quality and safety of maternal and newborn health services 15. Service Provider: A health worker qualified, employed and assigned tasks that provide a service to clients. 16. Skilled Birth Attendant: A doctor, midwife or nurse who has been trained to proficiency in the skills needed to manage normal labor and be able to identify, manage and refer mothers and newborns with complications with a goal of having a live baby and a healthy mother. 17. Standards: A set of behaviors or performance below which it is not acceptable ethically, professionally and technically. Standards are facts or consensus-based minimum requirements according to what is considered to be the norm. xi

EXECUTIVE SUMMARY This project report is about a quality improvement (QI) project which aimed at improving the quality of maternal delivery and newborn care services in Mityana hospital through increased availability and training of midwives. We designed the project using the Continuous Quality Improvement (CQI) approach which takes the CQI team through 7 seven steps of improving quality: (1) reasons for improvement, (2) current situation, (3) analysis, (4) countermeasures, (5) results, (6) standardization and (7) future plans. We focused our intervention on making structural review and system improvements in (staff availability, functional life-saving equipments, essential medicines and blood and support supervision). The Quality Improvement process 1. Establishing reasons for improvement: The project was initiated with a rapid assessment to establish the status of quality of maternal delivery and newborn care. The purpose of the assessment was to identify structural and process-related barriers to quality improvement. The assessment was carried out in January and February 2012. We interviewed 17 midwives, held 2 Focus Group Discussions, made some observations and reviewed data from maternity register. We used knowledge of health worker on Active Management of Third Stage of Labour (AMTSL) and use of partograph as key indicators for reducing maternal life- threatening risks such as severe anaemia (due to Post Partum Hemorrhage) and obstructed labour. We observed facility readiness to quality improvement by assessing availability of lifesaving drugs and other key medical supplies. We also held two Focus Group Discussion (FGDs) meetings with 20 clients (10 clients per group) who had received delivery services in the hospital. We reviewed data from maternity register for 2011 to obtain conditions related to maternal and neonatal morbidity and mortality including pregnancy outcomes. From interview of 17 midwives, 3 (17.5%) had ever heard about the term AMTSL: We asked midwives to describe the process of managing labour so that xii

we can determine inclusion of AMSTL and its components). Only 4 (23.5%) were able to mention all the 3 components of AMTSL. Response counts of specific AMTSL components were: (a) use of uterotonic (11 responses), controlled cord traction (8 responses), massaging uterus after delivery of placenta (9 responses) The hospital had adequate supply of oxytocics, and antibiotics and magnesium sulphate. Blood was frequently out of stock risking lives of mothers that needed it. Many partographs were incompletely plotted. Occasionally, waste disposal was delayed causing a stinking smell on the ward. The disposal was so crude that sharps, placentas and used cotton and gauze were dumped in uncovered pit; which is why vultures were common birds in the hospital. There was an infection prevention committee but it was non-functional. There was no running water at service points and resuscitation equipment was faulty. Clients were concerned about fewer midwives on the ward the extent that some delivered in absence of a midwife. No quality improvement team that had been formed. Clients were also concerned about waste management as they said that sometimes the smell from decomposing wastes in waste bins were more discomforting than labour pains. Clients complained of unfriendly response from some midwives especially at night. In case of referral, clients fended for themselves using expensive private means as there was no ambulance. The commonest risks associated with maternal morbidity were pre and postpartum hemorrhage, ruptured uterus, sepsis, obstructed labour and positive HIV status 2. Defining the current situation: In response to the rapid assessment results, a meeting was held at Mityana hospital drawing 28 top hospital and departmental leaders to intervene using CQI approach. A hospital CQI committee was established with a task to developing a one year roadmap for improving service delivery in maternity ward. Lessons that would be leant in the medium term would inform initiation of CQI projects in other departments. The CQI committee members reexamined the current system and process-related issues by listing, xiii

sorting, mult-voting and prioritizing problems that need intervention in order to improve quality of care in maternity. Fewer and less skilled midwives in maternity were the most voted problems that needed immediate attention. However; some of the gaps identified by the rapid assessment exercise were immediately addressed such as repairing resuscitation equipment and procurement of lanterns as an alternative source of light. 3. Analysis: We listed and mapped out where midwives were located in the hospital departments, in order to explore possibility of redeploying some of them to maternity. The established midwifery posts at the hospital were 25 but there were 38 midwives on the hospital staff list. Comparatively, of the 47 established s for nurses, only 29 were filled. This showed overstaffing of midwives. In spite of the excess midwives, the clients complained of inadequate number of midwives. A fishbone analysis was developed to find why there was low staffing of midwives yet the hospital had more midwives than they required. It was found out that several midwives were assigned duties on general wards in attempt to address the shortage of nurses leading to inadequate staffing in maternity ward The main reason why available midwives had inadequate knowledge and skills was that there were limited opportunities for refresher courses and even the few opportunities were not equitably distributed among all the midwives 4. Countermeasures: To address the poor quality of maternal and newborn services, Mityana hospital and Mildmay Uganda (an integrated HIV/AIDS implementing partner) jointly developed an intervention to address the quality gaps identified by the assessment exercise in maternity ward where over 5000 mothers are delivered per year. The fellow was tasked to guide the development of the implementation plan for the proposed intervention. The plan became the fellow s programmatic activity. xiv

The planned interventions were (1) review staff deployment and reallocate 6 midwives from other departments to maternity ward (2) conduct refresher training for all the midwives serving in maternity and (3) organize midwives to form CQI teams to progressively review and improve program performance. Expected results from interventions were improved labour monitoring through effective use of partograph. This would further reduce Fresh Stillbirth (FSB) because obstructed labour and fetal stress would be detected and addressed in time. Postpartum Hemorrhage (PPH) would reduce due to effective management of third stage of labour. Trends of maternal and newborn sepsis would decline due to improved infection prevention. These expected results would be augmented by improved equipment especially resuscitation machines, improved lighting and better waste management. Refresher training was organized with purpose to increase knowledge and skills among midwives to effective management of labour and minimize risk intrapartum factors. We used The Training Manual for Midwives in Provision of Integrated Reproductive Health of Ministry of Health (MoH). The topics selected included overview of reproductive health, effective management of labour, provider-client communication, quality improvement principles and practices, infection prevention and PMTCT. Two trainers were identified from the hospital i.e. the incharge maternity ward and the Principal Nursing Officer (PNO). Both were national trainers in integrated reproductive health care service. The fellow together with the trainers developed the training content based on training needs identified from the assessment exercise. Twenty four midwives were selected and trained. The training was divided into two 4-days training session and each session having 12 participants. Two more days per session were dedicated two hands-on practices on ward with support of a coach 5. Results: The redeployment review resulted in obtaining six more midwives who were deployed to maternity. It was not possible to get all 6 midwives from the hospital departments because there were few nurses in such departments. xv

Two were reallocated from the Antenatal Clinic and 4 were mobilized from other health facilities through the District Health Officer (DHO) with support from the Chief Administrative Officer (CAO). Two of external nurses and two others were midwives: The nurses were allocated to female ward and OPD which in turn released one midwife each. The two external midwives were absorbed in ANC clinic. Consequently, ANC released 2 midwives that became part of maternity ward staff. We trained 24 midwives: The training report revealed that midwives had an average score of 48.96% a (range of 31% to 67%) in a pretest; and average of 80.46% (range of 57% to 98%) in post test. After two months of redeployment and training of midwives, we did review of results: There was improvement in quality of maternal and newborn care based on these indicators: the level of completion of partographs per month was raised, there was a downward trend of maternal and neonatal sepsis, and FSB and postpartum hemorrhage reduced. Structural changes that are related to the intervention included high level of participation of staff, health managers and district leaders. The waste management has improved. The waste dumping ground has been relocated to a more secure place. Two resuscitation machines were repaired. The hospital procured lanterns to offer alternative light when there is electricity load shedding. 6. Standardization and future plans: Four CQI teams were formed in maternity ward and were being supported by hospital-level CQI committee to implement quality improvement projects. The projects are: (i) (ii) Reducing waiting time between maternity ward and theatre for mothers needing emergency obstetric care Improved inter-clinic referral of exposed babies to maternal HIV infection (maternity and Early Infant Diagnosis clinics) (iii) Infection prevention xvi

(iv) Adherence to standards of delivery and newborn care The hospital CQI committee is examining the possibility of scaling up CQI to other departments. However, it has been noted that some lessons from maternity are informing management of processes on other wards. There were challenges in getting the required number of midwives in maternity: some could not be removed from other stations because it could cause severe shortage; others had social and physical constraints that couldn t allow them serve in maternity. Several midwives and nurses asked for maternity leave at the same time as most them were in their early reproductive age and in need to have children. One lesson is that CQI process helps in identifying many quality gaps and with support of the health facility managers and local leaders, some quality gaps can be immediately addressed (such as staffing) and with support of partners, resources can be mobilized. We conclude that we succeeded in attaining short-term results such as redeployment and increasing knowledge of service providers. We identified that although the hospital was staffed with midwives beyond the staffing norms, shortage of nurses caused virtual low staffing of midwives. At process level, we see some changes after two months of reorganization of staff and training: pregnancy monitoring has improved as reflected by increasing trends of completely filled partographs. There is lowering trend of FSBs suggesting better pregnancy monitoring of labour that leads to early detection of fetal distress and action. In 11 months of 2012 the caesarian section rate was 22.4% compared to 24 % in 2012. However in 2012, there are lowering monthly trends in mothers who had normal vaginal deliveries compared with those delivered by caesarian section. This suggests that emergency obstetric care is prompt due to early detection of the need. We recommend regular staff deployment review to address internal staffing challenges that would not be seen at organizational level. Maternity staff needs xvii

regular refresher courses to maintain a high level of standard of care. Regular staff support supervision, regular check of functionality of delivery equipment and dissemination of service guidelines may improve quality consequently lowering maternal and neonatal morbidity and mortality. xviii

SECTION 1. INTRODUCTION AND BACKGROUND 1.1. Introduction Most life-threatening conditions of newborns and mothers occur during and after childbirth process are preventable. They include: sepsis, Postpartum Hemorrhage (PPH), eclampsia, Mother To Child Transmission (MTCT) of HIV and hypothermia (Mulumet et al, 2011; Kerber et al, 2007). To prevent and mitigate maternal and neonatal life threatening conditions, health systems are designed to increase Skilled Birth Attendance (MoH, 2007). Community mobilization programs encourage health facility-based deliveries with assumption that it is safer to deliver in health facilities than at home. However, the institutional quality of maternal delivery and newborn care services by skilled attendants remains poorly rated especially in developing countries and thus leading to persistently high maternal and neonatal morbidity and mortality (Mulumet et al, 2011; Van de Broek et al 2009;ICM, FIGO, WHO, 2006). A combination of a SBA, appropriate medicines, equipment and infrastructure provide a safe and clean delivery environment in which mothers and newborns can survive the life threatening conditions (Bhutta et al 2010). Quality of maternal and neonatal care must continuously improve to reduce the vulnerability of pregnant mothers and newborns to life-threatening conditions. Quality improvement is systematic, data-guided activities that are specifically designed to cause prompt and substantial improvements in performance of health process (USAID, 2012). According to Hulton et al (2000), QI in maternal health care is the degree to which maternal health services for individuals and populations increase the likelihood of timely and appropriate treatment for the purpose of achieving desired outcomes that are both consistent with current professional knowledge and uphold basic reproductive rights. It is never an accident but a result of high intention, intelligent direction and skillful execution of appropriate interventions; and systematic implementation to reduce health risks in health facilities (Deboult & Mallen 2012; Massaoud et al, 2002). 1

Quality improvement approach ensures that standard guidelines which midwives and other skilled birth attendants will use are available and utilized. Inadequate, ill-equipped and unskilled SBA contribute to poor quality of maternal and neonatal care services; for example, Prabhath et al (2002) found out that 73% of maternal deaths that occurred in tertiary health facilities in the developing world were due to substandard care offered by SBAs. 1.2. Trends in maternal and neonatal morbidity and mortality The global burden of maternal and newborn morbidity and mortality is high: out of global 536,000 maternal deaths each year, 99% are from developing countries; 80% of the deaths are preventable with timely interventions that are proven to be effective (Van de Broek et al, 2009). Annually, over 4 million neonates die worldwide (Mash et al, 2002; Lawn et al, 2009). The African region has the highest rates of neonatal mortality in the world but two thirds of these deaths can be averted with known strategies; however, this region has shown the slowest progress so far in reducing neonatal deaths (WHO, 2012). In Sub-Saharan Africa (SSA) alone, 279,000 neonates and 4.5 million infants die every year (Mwaikambo, 2010). The total loss of mothers, newborns and children in SSA every year is 4.7 million lives. Similarly over 880,000 babies are stillborn in SSA every year (Kinney et al 2010). In Uganda, maternal mortality ratio stands at 438 deaths per 100, 000 live births, and neonatal mortality rate at 27 deaths per 1000 live births and infant mortality at 54 per 1000 live births. The under five deaths has reduced from 143 to 90 deaths per 1000 live births between 2006 and 2011 (MOH, 2007; UBOS & ICF Int. 2011). Other studies show that in Uganda, 6,500-13,000 women and girls die every year leaving over 405,000 with chronic and debilitating effects (Futures Group, 2012). 2

Institutional delay in maternal delivery is one of the three delays that increase the risk to life loss among newborns and mothers (MoH, 2007). Other delays are: delay to decide to go for skilled birth attendance and delay to reach a health facility that has the capacity to meet emergency obstetric and newborn care needs of mothers and babies. Most of Maternal-To Child Transmission (MTCT) (60-70%) occur at the time of childbirth (Esiru, 2008). This was also documented by Wabwire-Mangen et al, 2008. This report focuses on Mityana hospital in Mityana district in Uganda where quality maternal and newborn care improvement project was implemented. Mityana Hospital is a public hospital with 100 bed-capacity. Every year, about 5,000 mothers are delivered in maternity ward of Mityana hospital. The ward also handles pregnant mothers with medical conditions such as malaria, those who have just delivered either by caesarian section or normal vaginal delivery and waiting mothers who have risk factors that need to be closely monitored towards labour. Mildmay Uganda and Makerere University School of Public Health-CDC (MakSPH-CDC) Fellowship program supported Mityana hospital to review and improve the quality of maternal delivery and newborn care services offered to clients that come to this hospital. An assessment to establish status of quality of delivery and newborn care in Mityana hospital was conducted in January 2012. Midwives were interviewed, Focus Group Discussions were held with clients and hospital data reviewed. Some key informant interviews were held and observations made and recorded by the assessing team. Results showed that there were several quality gaps: less than a quarter of midwives had comprehensive knowledge about Active Management of Third Stage of Labour (AMTSL). Although 38 midwives are employed in the hospital (13 in excess of established of 25 midwives), only 18 conducted deliveries and the rest were multi tasking on wards and clinics in the hospital. Midwives deployed in maternity are few compared to the daily workloads on the ward. Maternity waste and placenta were 3

mixed up and dumped in an open pit and some staff members were demotivated due to low pay. Power load-shedding was rampant; phone torches were usually used in delivery as alternative light since the cost of fuel for generator unaffordable by the hospital. Main transport for obstetric emergency was by private motorcycles. There were few functional resuscitation gargets: the manual sanction machine was faulty, the electric one lacked some parts. Clients complained of smelly wastes on ward and staff negligence. The common risk factors found from maternity register (2011) are shown in the figure below: - 25 20 15 10 5 0 Frequency of risk conditions among mothers and their babies in Mityana hospital in 2011 (%) n=116 20.7 19 17.2 8.6 7.8 7.8 7.8 6 5.2 Figure 1-1: Risk factors found among mothers attending maternity ward in Mityana hospital, 2011 (Mildmay Uganda, 2012) It is against this background that we purposed and pursued improvements in delivery improved quality maternal delivery and newborn care services using CQI approach as described in the following sections 4

SECTION 2. THE QUALITY IMPROVEMENT PROCESS 2.1. Reasons for improvement We drew form the assessment results to design a quality improvement project. The objective of improvement was to increase client access to effective skilled birth attendance in order to reduce occurrence of maternal life-threatening factors such as PPH, sepsis, obstructed labour; and neonatal health risks such as asphyxia intrapartum HIV transmission, sepsis and deaths within 24 hours. This would be achieved by identifying and reducing barriers to skilled care 2.2. Examining the current situation We held quality improvement consultative meeting that attracted 28 participants from various hospital departments to brainstorm on key causes of poor quality of maternal delivery and newborn care. A brainstorming session identified 17 problem areas. Discussants set criteria for sorting one priority problem to solve. The criteria were that the problem must be process-related, with locally available solution, with greater impact on reduction of maternal and newborn morbidity and mortality. Sorting left 7 key problem areas. Members held a 2-level multi voting: the first round used nominal scale approach where each member was allowed to vote only 3 problems and ranking them using a scale of 3:2:1. The most serious problem was given 3 scores, the next serious problem given 2 scores and the least important one 1 score. In round 2 problems that scored above 10 scores were selected. Members were requested to rank the remaining problem with the most important problem first and less important. Low staffing in maternity ward and limited skills among health workers were priority problems to solve because improvements in staff skills and staff numbers would form a firm foundation from which other CQI project will develop. 5

Table 2.1 shows how we arrived at the priority problem: - Table 2-1: Ranking of selected problems associated with poor quality of maternal delivery and newborn care in Mityana hospital. Problem Round 1 of voting (using nominal scale 3:2:1) Round 2 of voting (rearranging the first 4 problems in descending order of their strengths) Most midwives have inadequate skills Lack of supplies e.g. blood Mothers delay to come to deliver Low staffing in maternity ward 34 2 32 3 24 4 24 1 No resuscitation corner 9 Low staff morale 8 Poor waste disposal 7 A mapping exercise for midwives was done to locate where they are located. The table below shows their distribution:- 6

Table 2-2: Distribution of midwives in wards and departments of Mityana hospital Shift Duty No. Total On duty Day 3 8 Evening 2 Night 3 Off-duty After day-duty 3 8 Other assignments After evening-duty 2 After Night-duty 3 ANC clinic 7 20 Male ward 2 Paediatric Ward 2 Female ward 3 Out-Patient Department(OPD) 2 Chronic care clinic 2 Night superintendants 2 On leave 2 2 Total 38 Over 50% of midwives had no opportunity to conduct deliveries as they stayed in other departments rather than maternity for a long time. 2.3. Problem analysis To identify the root causes of low staffing and inadequate skills among midwives in maternity, a fishbone approach was used. Key roots were: inadequate opportunities for staff to acquire and improve on knowledge and skills in conducting standard delivery and newborn care. The figure below shows cause-causes of low staffing and less skilled midwives: - 7

Limited opportunities for staff to improve knowledge and skills in delivery and newborn care Lack of refresher training Available training opportunity dominated by a few midwives Inadequate/irregular/lack of effective support supervision Some disallowed to practice delivery as a disciplinary action. Few and less skilled midwives in maternity ward Exemption to deliver on medical grounds Some midwives lost interest to deliver mothers due to heavy workload. Few midwives available for maternity services Long time without staff reshuffle (over 6 months) Some midwives attracted to other clinics and wards due to shortage of nurses or dislike of difficult night duty associated with maternity Figure 2-1: Fishbone analysis showing factors associated with deployment of fewer and less skilled 8

The main cause of limited skills among midwives was inadequate opportunities for training: few training opportunities are dominated by selected staff loyal to decision makers. Some staff do not gain experience from practice because they are not regularly supervised or transferred to other areas that promote professional skill enhancement. The reason why midwives may be few in a ward could be caused by low staffing level of other health cadres leaving health managers with no option but to redeploy midwives to perform non-midwifery duties. Some midwives may dislike or fail to work in maternity due to various reasons such as health status of the midwife or availability of less stressful alternative duties. Some midwives overstay in midwifery-related stations but which are not involving delivery. This may lead to skill decline and poor performance 9

SECTION 3. PLANNING COUNTERMEASURES In response to the assessment finds and in pursuit of implementation of the roadmap for accelerating the reduction of maternal and neonatal mortality and morbidity in Uganda (MoH, 2007) we set out to improve maternal and newborn care services. 3.1. Intervention objective The proposed intervention was to improve the quality of maternal delivery and newborn care services through redeployment and capacity building of existing health workers in Mityana hospital. 3.2. Intervention strategies We set out to redeploy and train midwives as capacity improvement strategy to improve the quality of maternal and neonatal care services. Below were the intervention objectives: - a) Review midwives deployment strategy to raise number of midwives deployed to maternity ward of Mityana hospital from 18 to 24. b) Conduct a refresher course for midwives focusing on knowledge and skill gaps identified by the assessment report. c) Facilitate establishment of CQI teams in maternity ward to continuously review and improve maternal and newborn care 3.3. Expected results Key process results expected were that 6 midwives would be redeployed form other departments to maternity, 24 from maternity ward would be trained and consequently, trends in maternal and neonatal sepsis and PPH would reduce. 10

The indicators of success included improved labour monitoring. It was expected that as skills improve and workload reduce due to increased staffing in maternity, there would be more effective pregnancy monitoring with complete partographs. Improved monitoring of labour would result into early detection of fetal and maternal distress with timely management leading reduction in fresh stillbirths and other complications. Reduction of trends in maternal and neonatal sepsis would show improved infection prevention on the ward (improved practice). Another expectation of quality improvement would be compliancy to conducting AMTSL according to standards. Since the registers do not capture data on effectiveness of AMTSL, we targeted seeing reductions in PPH as an indicator of effective AMTSL the condition of PPH is the commonest cause of maternal mortality. 3.4. Implementation framework The implementation of the countermeasures is described in the table below: - 11

Table 3-1: Implementation framework Quality gap Countermeasure Indicator of improvement Baseline Target MOV Assumptions Few midwives in Redeploy more midwives No. of midwives 18 24 Minutes Redeployment won t cause maternity ward from other departments to redeployed and duty severe staff shortage at maternity roster departmental level Inadequate knowledge & Train all midwives No. of midwives 0 24 Training All midwives planned to be skills about delivery and newborn care among midwives working in maternity deployed in maternity to deliver mothers and care for babies according to standards trained Knowledge gain by participants unknown post test with at least 65% report Pre and post test results deployed to maternity will be available for training No competing priority programs at the hospital Midwives unable to offer Trained midwives able to No. of unknown 80% Data from Midwives are continuously standard care deliver and care for partographs maternity supervised and supplied with mother and baby completely partographs, medicines and according to standards filled other supplies Persistently raising trends Maintain a standard of Reducing trends Data Not Lowering Midwives posted in maternity of key risk factors care using knowledge & of PPH, sepsis, from known trends are not immediately transferred skills from training fresh still births, maternity after register training 12

3.5. Reviewing staff redeployment The fellow engaged the in-charge of maternity ward, the Principal Nursing Officer and the Medical Superintendent in dialogue to discuss the deployment strategy of midwives so that more will be added on those currently allocated to maternity without compromising the functions of other departments which are supported by midwives. The hospital senior managers undertook the task of reviewing the staff deployment strategy to identify 6 more midwives to redeploy in maternity. It was found out that the staffing norm for midwives in the hospital was 25 but the staff list had 38. Comparatively the staffing norm for nurses was 47 but there were only 29 nurses. When adjustments were made to find which midwives to post to maternity, only two from ANC clinic were identified. A series of dialogue meetings with district leaders and hospital stakeholders lead to securing 4 nurses and 2 midwives from other facilities that made it possible to have 6 midwives redeployed in maternity. 3.6. Training Midwives at Mityana hospital Twenty four midwives were identified for training with the view that they will take turns to attend to mothers and babies in maternity. Training was preceded by preparatory training meetings at Mityana hospital. We developed the training content based on Integrated Reproductive Health Training Manuals of MoH. We focused the training on the overview of reproductive health services, Focused ANC, principles of CQI, managing labour with emphasis to 3 rd stage and use of partograph. Other topics covered were communication skills, emergency situations for mothers and neonates, infection prevention and introduction to PMTCT. The refresher course was divided into two 4-days session; each session had 12 trainees. Each group of trainees had two more days for practicing with a coach 13

on ward. We evaluated trainees with a pre and post test followed by observations as they did practical session under supervision of a trainer as a coach. 14

SECTION 4. THE QUALITY IMPROVEMENT OUTCOMES 4.1. Redeployment of midwives We had planned to adjust staff redeployment and redeploy 6 midwives from other departments to maternity so that staffing in maternity ward can increase from 18 to 24 midwives. We mapped location of all 38 midwives considered work schedule (morning, evening and night duties) and workloads in various departments. To get six more midwives to deploy to maternity, we attempted to get all midwives from other departments but we succeeded to get only 2 from ANC clinic. We could not redeploy all the required number of midwives from the hospital departments because it would cause severe staff shortage at department level. Engaging both hospital and district stakeholders (as individuals and groups) through a series of meetings led to deploying 4 more staff (2 midwives and 2 nurses) from other health facilities in the district to the hospital. The two external nurses went to female ward and OPD which in turn released one midwife each. The two external midwives were absorbed in ANC so that 2 more midwives (from existing staff) were redeployed to maternity ward. The final adjustments led to staffing in maternity raise from 18 to 24 midwives achieving 100% of targeted staffing in maternity ward. 15

Figure 4-2: Participants of the stakeholders Meeting Figure 4-1: At the stakeholders meeting; the fellow makes a point as the RDC listens 4.2. Training process In September 2012, we trained of 24 midwives. The training was organized in two phases and each phase taking 12 trainees. We used 4 days to provide theory and two days for practice. While the first group was starting their practice on 5 th day, the second group had their first day of theory study. This led to running concurrently the practical session for the first group of trainees with theory of the second one. This provided trainees with 6 days of training both in class and on the ward. Generally, the trainees performed well as reflected in the knowledge acquisition. The combined results of the pre and post test below show how trainees acquired knowledge (1 st group is no.1 to 12 and 2 nd group is no. 13 to 24): - 16

% Trainee identification No Figure 4-3: Pre and post test results for 24 midwives trained at Mityana hospital. The trainees had an average score of 48.96% a (range of 31% to 67%) in a pretest; and average of 80.46% (range of 57% to 98%) in post test. 4.3. Other changes The hospital management first addressed some gaps that were identified during quality assessment: these include, repairing recitation machines, getting alternative light to address lighting challenges when electricity load shedding takes place. These were fixed: 4 lanterns were bought, 2 sanctions machines were repaired and the hospital did not renew the contract of the supplier of sanitation services because he had not managed to support staff to improve waste management. 17

The following are some of the photos taken during the training sessions: - Figure 4-5: One of the groups of trainees doing 5 pre-test Figure 4-4: Midwives and trainers ready for a practical session after 4- days theory Figure 4-7: A midwife assisting a mother To breastfeed as the fellow observes 4.4. Intervention outcomes Figure 4-6: A midwife gives a health talk mothers at discharge after delivery by C/S We expected change behavior of health workers conducting delivery such as plotting and completing partograph we targeted improvements in effective 18

AMSTL. We did not collect data on effectiveness of AMTSL since it was not captured by data from maternity register. We expected that reductions in PPH would show how midwives were effective in conducting AMTSL The figure below shows trends in completion of partographs: - Partograph completion initiated (July 2012) No. Training takes place (Sept. 2012) Months Figure 4-8: Trends of completely plotted partographs (January to November 2012) In July, midwives were encouraged to complete partographs as a measuere to demonstrate that labour was completely monitored. The trends of completely partographs increased from July and peaked in October, one mothe after training midwives. We observed rising trends in some of the indicators of quality improvement outcomes: Below is a figure showing trends of key risk factors associated with 19

poor quality of maternal delivery and newborn care: NO. Months Figure 4-9: Trends of maternal and neonatal sepsis, postpartum hemorrhage and Fresh Stillbirth (FSBs) 20

The trends in the risk factors were higher in September (when training was done and sharply lowered in October. There is lowering trend of FSB after September training. It is not clear why most of the trends increased during the month of training (September) but the reductions in the following month suggest quality could have improved. No. Figure 4-10: Trends in delivery methods and FSBs in relation to completely filled partgraphs. The raising trend of complete partographs followed by flatening trends of FSB, the raising of monthly cases of mothers who delivered by caeseian section and the lowering of number of mothers had biths suggest improvements of pregnancy outcomes through completion of partographs 4.5. Qualitative evaluation Before the pretest, trainees could not believe that they had such knowledge gap. One trainee stated: We did not know how uninformed we were until you showed us the pretest results! Months Several trainees said they had been giving oxytocin to mothers during labour to minimize hemorrhage after birth but could not mind about injecting it in 21

stipulated time. One midwife stated: I knew that giving oxytocin injection was to prevent severe postpartum hemorrhage but I did not recognize the importance of giving it as soon as the baby is delivered. Some midwives had negative attitude about filling a partograg. One of the senior midwives noted: Before this training, I disliked filling the partograph because it looked so complicated and I felt that I could still detect danger without it; I now feel that filling a partograph is important and not so hard. Another midwife felt that she had negative attitude about clients conduct: I thought clients were to blame for most problems during labour and handling of babies; now I feel we health workers fail to do some things or do them haphazardly because of lack of knowledge or what to use. This causes poor quality of services. Some midwives did not know that their supervisors were professional trainers as noted by one midwife: We are lucky that our supervisors are also trainers; they can help us do better if they planned to coach us on ward on regular basis 4.6. Institutionalization of CQI After the training, the hospital management used experience from maternity to design scale-up strategy to enable other departments also improve the quality of service they offer to clients. At the hospital level, an interdepartmental CQI team was formed to oversee improvement initiatives in all departments. The maternity staff continued CQI by identifying more QI areas. All 24 trained midwives grouped themselves in 4 teams each targeting a specific thematic area. The themes of focus were: reduction of number of babies lost to follow-up between maternity ward and EID clinic, infection prevention, adherence to delivery standards, improving client flow to reduce waiting time and improving provider-client relations. 22

During the process of discussing CQI, staff members were pleased to participate in program planning as stated by one junior midwife: We used to wait for decisions from our seniors, but with quality improvement process everyone participated. It was noted that for quality to improve, interdepartmental linkages must be strengthened. One senior midwife commented that: We need to improve coordination with theater staff because some deaths occur due to delayed response from there. We should educate mothers not to delay, and health workers in lower units should be told to refer mothers in time. Another midwife commented: I appreciate the part of using data to show how well or poor we are performing The fellow observed that like any other skill-based task holders, midwives need routine refresher courses at the facility emphasizing hands on sessions. More that 60% of midwives had difficulties in plotting a partograph; actually most of the mothers were delivered without it. The understanding of common concepts among midwives was still low; for example 15 out of 24 midwives did not know the meaning of nosocomial infections and some who attempted called them infections of the nose instead of infections acquired by patients/clients from the hospital. The universal precautions concept (that outlines infection prevention key practices) was not known by all the groups. For example when the groups were made to discuss this package, only 3 out of eight precaution measures for infection prevention came out. During practice on the ward, most of the procedures were correctly done except resuscitation of the baby using ambubag and cardiac massage. The concept of CQI was new to almost 90% of participants. In conclusion the programmatic activity was successfully completed. We planned to deploy 24 midwives and now they are on duty roster of maternity ward. We had planned to train 24 midwives and all were trained. What remains 23

is trainee follow-up and mentoring CQI teams to effectively manage their projects and document final quality status. We did not conduct client consultation to measure change in satisfaction. The project that was planned to start in May actually started in August. This was due to delayed process of redeployment and postponement of training midwives by district leaders due to competing priorities such as national HPV vaccination campaign. 24

SECTION 5. FUTURE PLANS The implementation of the programmatic activity attracted health workers who appreciated the CQI approach to program performance improvement. The hospital management planned to continue more CQI projects in maternity as other departments learn so that lessons learnt can be scaled up to other hospital departments. One CQI team was formed to plan quality improvements at organization level. In maternity ward, 4 sub-teams were established to create CQI projects on 4 thematic areas: one for improving inter-clinic referral of exposed babies to maternal HIV for Early Infant Diagnosis (EID), improving infection prevention, reduction of waiting time, improving customer communication and improving standard of delivery and newborn care. Each team would brainstorm on quality-related problems in each thematic area prioritize key problems, develop intervention plans, monitor and communicate results during departmental meetings and Continuing Medical Education sessions. Immediate results would be charts of standard operating procedures, review meetings held and copies of work plan in place. Medium term results would be performance review charts, improvement in client satisfaction, reduced incidence of severe hemorrhage and sepsis. Long range results would be reduced morbidity and mortality of mothers and babies, reduced maternal transmission of HIV to exposed babies and reduced complaints work overload among staff members. By the time of developing this report, teams had reached brainstorming stage of CQI process. 25