Introduction

The World Health Organisation (WHO) defines postpartum haemorrhage (PPH) as vaginal bleeding in excess of 500ml after childbirth [₁]. Globally, PPH occurs in 10.5% of live births [₂]. It is the leading cause of maternal mortality worldwide (25%) and a major contributor of maternal deaths in Africa (33.9%) [₃,₄]. Factors associated with PPH include increasing maternal age, primiparity, foetal macrocomia, multiple pregnancies, fibroids, antepartum haemorrhage, history of PPH, previous Caesarean section, prolonged labour, and episiotomy [₅,₆]. However, none of these factors has an adequate positive predictive value for a good screening tool. Therefore, experts recommend that all women should benefit from active management of the third stage of labour, the only intervention known to prevent PPH [₁]. Active management of the third stage of labour reduces the incidence of PPH by one third, and therefore a significant proportion of women will proceed to develop PPH despite this active management [₇]. Skilled birth attendants most therefore be conversant with proper management of PPH.

In Malawi, 43% of women give birth outside health facilities [₈], implying that these women do not have access to timely and quality emergency obstetric care. When PPH occurs outside the health facilities, a significant proportion of women will die before arrival to the health facility. If they are fortunate to arrive to a health facility they are usually in a state of shock with very low blood pressure, weak and fast pulse, cold and clammy skin, rapid breathing, anxious or confused, and sometimes unconscious. Many women will go into shock with moderate bleeding because the baseline haemoglobin is usually low [₉]. Therefore in health facilities, birth attendants should take extra care and respond rapidly to any signs of moderate to excessive bleeding or shock following childbirth. A confidential enquiry into institutional maternal death in Malawi revealed that obstetric haemorrhage is the cause of 26.7% of maternal deaths that occur within 24 hours of admission and identified poor quality of care as a major contributing factor [₁₀]. This enquiry recommended the use of criterion based audit to improve the quality of emergency obstetric care in Malawi.

We sought to assess and improve the management of postpartum haemorrhage in Malawi by introducing criterion based audit in three districts.

Methods

Study setting

Clinical audit cycle

The classic steps of a clinical audit cycle were followed in this study (Figure 1). The first audit was conducted in May and June 2007 and the second audit in October and November 2007. In-between the two audits, we allowed a period of three months to implement the changes.

Figure 1

Figure 1: Clinical audit cycle

Step 1: Establishment of standards for postpartum haemorrhage

We used evidence from existing guidelines, namely Malawi national guidelines and WHO manuals [₁₂,₁₃]. Where necessary this was supplemented by evidence from the Cochrane database, standard textbooks and articles from peer-reviewed journals. A multidisciplinary team (including maternal health experts and policy makers) established local standards for obstructed labour, during a workshop that brought participants from the three districts. At the end of this two day workshop, the participants agreed on 10 objectives and developed the structure, process and outcome criteria for each objective. The initial list of 205 criteria was reduced to a shorter list of 6 criteria by identifying the most important criteria for the management of PPH. Postpartum haemorrhage was defined as showed in Table 1. The six criteria for the management of PPH are presented in Table 2.

Figure 2

Table 1: Working definition of postpartum haemorrhage

Figure 3

Table 2: Standards for the management of postpartum haemorrhage

Step 2: Measurement of current practice

Current practice was measured by a retrospective review of case notes. A multidisciplinary quality improvement team (made up of nurses, midwives, laboratory technicians, clinical officers, and doctors) was established in each hospital. The teams reviewed the cases of PPH in May and June 2007 and extracted the data into a data collection sheet. Forty cases of PPH were reviewed in the 8 hospitals.

Step 3: Analysis of findings and feedback

The percentage of attainment of each standard was calculated and the results presented during a quality improvement workshop that brought together representatives from the 8 hospitals.

Step 4: Recommendation and implementation of change

The gaps between current level of care and standards were discussed during the quality improvement workshop (see step 3). Most recommendations made were common to all hospitals, although a few recommendations were specific to some hospitals. When the participants returned to their hospitals, they briefed the rest of the staff and management about the gaps identified and the recommendations made during the workshop. The staff and management then worked jointly to implement the recommendations. The recommendations made and implemented are presented in Table 3.

Step 5: A re-audit of standards to assess progress

Three months later after the first audit, a second audit was performed to assess the progress made. Forty five case notes were reviewed during this second audit.

Figure 4

Table 3: Recommendations made and implemented after the initial audit

Statistical analyses

Results

Baseline characteristics

The mean age was 25 years in the initial audit and 24 years in the re-audit, but there was no statistically significant difference in age between the audits (p = 0.646) (Table 4). Similarly, median parity and percentage of primiparous women were similar in the two audits. There was no significant change in perinatal mortality (p = 1.000). However, there was a trend towards a decrease in case fatality rate (10.0% vs 6.7%).

Figure 5

Table 4: Baseline characteristics

Audit results

When the first audit was performed, wide gaps were identified between current practice and four standards: typing and cross-match carried out (65.0%), patients' haemoglobin or hematocrit carried out (67.5%), vital signs monitored closely (32.5%), and fluid intake/output chart set up (0.0%). Two standards were attained in 100% of cases: intravenous (IV) access and continuous administration of IV fluid until cross-match blood was available, and administration of oxytocic drugs.

When the second audit was conducted, there was a significant improvement in three standards: typing and cross-match carried out (p = 0.034), patient's hematocrit or haemoglobin established (p = 0.029), and fluid intake/output chart maintained (p <0.001). There was a trend towards an improvement in the close monitoring of vital signs (32.5% vs 53.3%). There was no significant change in two standards in which attainment was 100% during the first audit: IV line set up and IV fluids given (100.0% vs 97.8%), and oxytocic drugs administered (100.0% vs 95.6%).

Figure 6

Table 5: Attainment of standards in the first and second audits

Discussion

This study describes the use of criterion based audit to assess and improve the management of PPH in Malawi. There were significant improvements in some aspects of PPH management, namely typing and cross-match, carrying out patient's hematocrit or haemoglobin, maintaining fluid intake/output chart and a trend towards an improvement in the monitoring of vital signs. However, there was no significant change in the administration of IV fluids and oxytocic drugs, because almost all women with PPH in both audits receive IV fluids and oxytocic drugs.

Before the introduction of criterion based audit in the three districts, typing and cross-match, establishing of patient's haemoglobin or hematocrit, close monitoring of vital signs and maintaining a fluid intake/output chart were not carried out routinely in all patients with PPH. They were only done at the discretion of the clinical officer or doctor taking care of the patient. Although many cases of PPH treated in hospitals occur after hospital delivery, some are referred from health centres or from traditional birth attendants. Referred cases often arrive the hospital in a state of shock and many did not recover from the shock. The introduction of criterion based audit helped standardise and improve the management of PPH in the three districts.

Criterion based audit has just recently been introduced in developing countries [₁₄,₁₅,₁₆]. Earlier studies assessed the feasibility of criterion based audit maternity units in Uganda [₁₄], and Ghana and Jamaica [₁₅,₁₆]. Similar to findings from previous studies, we found that criterion based audit was feasible in resource-limited countries. However, the way we established standards for PPH differed from traditional method described by other authors [₁₄,₁₅,₁₆]. Traditionally, standards have been developed by a panel of experts and implemented by a multidisciplinary team. In obstetrics, this panel of experts is usually made up of obstetricians, while midwives only get involved during the implementation phase. We involved midwives from the very process of developing standards and this promoted ownership and sustainability among the health care providers. We equally involved policymakers and managers from the early stages of criterion based audit and this facilitated organisational changes, release of finances and resources, and approval where necessary to help implement the changes recommended by the audit.

We encountered some challenges in the process of introducing criterion based audit in Malawi. These challenges included shortage of staff, difficulty changing old practice, inadequate knowledge and skills to manage emergency obstetric complications, poor documentation, and lack of resources. Fortunately, all these challenges except shortage of staff were overcome by the very process of criterion based audit. Three months between the initial audit and the re-audit was too short for us to expect a significant change in staffing level. However, criterion based audit helped improve documentation especially cases notes and maternity registers, and devised ways of introducing new practice without making the providers feel threatened [₁₇]. Quality improvement is generally believed to be a very expensive process. Fortunately, criterion based audit is an inexpensive tool. Therefore, using criterion based audit we were able to implement effective and locally appropriate recommendations using very few resources. The knowledge and skills of providers were improved by delivering a three day life saving skills – emergency obstetric care course designed and delivered by the Royal College of Obstetrics and Gynaecology International Office [₁₈]. Two courses have been delivered in two districts (Kasungu and Salima), while another course is planned for the third district (Lilongwe). Comparing knowledge before and after the course, we found a significant improvement in knowledge on the management of obstetric emergency complications including PPH.

In conclusion, criterion based audit can improve the management of PPH countries with limited resources. We suggest that standards with 100% level of attainment during the initial audit should be re-audited because a drop in this level could be picked up by the re-audit.