Improving the quality of obstetric care for women with obstructed labour in the national referral hospital in Uganda: Lessons learnt from criteria based audit

Background: Obstructed labour remains a major cause of maternal morbidity and mortality whose complications can be reduced with improved quality of obstetric care. The objective was to assess whether criteria-based audit improves quality of obstetric care provided to women with obstructed labour in Mulago hospital, Uganda. Methods: Using criteria-based audit, management of obstructed labour was analyzed prospectively in two audits. Six standards of care were compared. An initial audit of 180 patients was conducted in September/October 2013. The Audit results were shared with key stakeholders. Gaps in patient management were identified and recommendations for improving obstetric care initiated. Six standards of care (intravenous fluids, intravenous antibiotics, monitoring of maternal vital signs, bladder catheterization, delivery within two hours, and blood grouping and cross matching) were implemented. A re-audit of 180 patients with obstructed labour was conducted four months later to evaluate the impact of these recommendations. The results of the two audits were compared. In-depth interviews and focus group discussions were conducted among healthcare providers to identify factors that could have influenced the audit results. Results: There was improvement in two standards of care (intravenous fluids and intravenous antibiotic administration) 58.9 % vs. 86.1 %; p < 0.001 and 21.7 % vs. 50.5 %; P < 0.001 respectively after the second audit. There was no improvement in vital sign monitoring, delivery within two hours or blood grouping and cross matching. There was a decline in bladder catheterization (94 % vs. 68.9 %; p < 0.001. The overall mean care score in the first and second audits was 55.1 and 48.2 % respectively, p = 0.19. Healthcare factors (negative attitude, low numbers, poor team work, low motivation), facility factors (poor supervision, stock-outs of essential supplies, absence of protocols) and patient factors (high patient load, poor compliance to instructions) contributed to poor quality of care. Conclusion: Introduction of criteria based audit in the management of obstructed labour led to measurable improvements in only two out of six standards of care. The extent to which criteria based audit may improve quality of obstetric care depends on having basic effective healthcare systems in place. This study was conducted in the obstetric wards in Mulago Hospital, Uganda’s National Referral Hospital and teaching hospital for Makerere University College of Health Sciences. Most of the patients are referred from satellite health facilities around Kampala City. Other patients are referred from Regional referral hospitals over 250 miles from Mulago hospital. Mulago hospital is a government-funded 2700-bed hospital, which currently serves more patients than 60,000 obstetric patients per annum. The hospital has 33,000 deliveries per year. The care is free to the public. The labour ward receives about 60–80 admissions, delivers about 50–70 mothers by vaginal delivery and 20–25 mothers by emergency caesarean delivery daily in 12-h duty shifts. Most of patients get spinal anesthesia. General anesthesia is only given in cases were spinal anesthesia fails to take or incase of obstetric complications such as HELLP syndrome. Operative vaginal delivery is less than 10 % at this facility. The study units included the labour ward, postnatal ward, Operating theatre and HDU, which units are adjacent to each other on the same floor of the hospital. The medical care is provided via a 24-h duty shift team comprised of two specialist obstetricians (as team leaders), seven residents (trainee obstetricians), three Intern doctors, eight midwives, one records officer and three Anesthetic Officers / Anesthesiologist. This translates into provider: patient ratio of 1:5. The study design was a criteria-based audit conducted prospectively over a period of seven months. Two criteria-based audits were performed, one from September – October 2013 and the second from February – March 2014. In the interim between the two audits, a quality improvement initiative was held in November 2013 when providers and stakeholders were presented with the results from the first audit. The initial audit results were discussed with 53 key stakeholders (specialists, resident obstetricians, intern doctors, midwives and hospital administrators). Gaps in patient management were identified and recommendations for improving obstetric care agreed upon for implementation. Six standards of care (intravenous fluids, intravenous antibiotics, monitoring of maternal vital signs, bladder catheterization, delivery within two hours, and blood grouping and cross matching) were implemented (Tables 1 and ​and2).2). Management protocols for obstructed labour including the six standards were developed and implementation was initiated after training of all healthcare providers. Observing criteria-based audit, a review process was conducted where clinicians were to agree on a number of explicit and realistic criteria of good quality for mothers with obstructed labour that were feasible within Mulago hospital [5, 6]. In absence of local protocols and guidelines on management obstructed labour in Mulago Hospital, in this study we used a criteria (standard of care) suggested by Graham et al. with some modifications [7]. A second audit was conducted four months later to evaluate the impact of implementing the recommendations, and results of the two audits were compared and a mean care score computed. In-depth interviews and focus group discussions were conducted to assess factors that could have influenced the results. The steps that were followed in this study are demonstrated in Fig. 1. Indicators, their set targets and Action points following presentation of first audit results Indicators, their set targets and Action points following presentation of first audit results Ω Labour suite laboratory to process all the samples; Doctors and midwives to take off samples; Grouping and cross matching may be done in antenatal care to reduce delays Shows the steps that were followed in the Audit process at Mulago Hospital Upon making a diagnosis of obstructed labour by the obstetric team on duty, patients who consented to participate in the study were followed prospectively to the point of discharge from the facility. The quality of obstetric care given to them was analyzed from their charts. Participants were interviewed for details where there was need for clarity in cases of under or over documentation of clinical findings. Inclusion criteria: Exclusion criteria: The study participants were 180 women with obstructed labour in each audit that was conducted within 24 h of admission. The criteria used for diagnosis of obstructed labour included having three adequate uterine contractions in ten minutes lasting 45–60 s with no cervical progression in two hours or progression less than 1 cm/h or presence of clinical signs of obstructed labour (impending uterine rupture, Bandl’s ring, blood-stained urine, severe caput or molding of fetal head) or clinical signs of shock (hypotension or tachycardia with a pulse of >110/min). The maternal outcomes, which were considered in the study included; hospital stay of the participants, mode of delivery, maternal morbidities like puerperal sepsis, uterine rupture, fistula formation, and whether mother died or was alive at discharge. Fetal outcomes that were considered included; admission to a NICU (special care Unit), rationale for admission to the NICU, and whether the baby died or was alive at discharge from hospital. An initial one-week pilot study of ten participants diagnosed and managed for obstructed labour was conducted to pretest the instruments. After revising the instruments, audit participants were recruited from the labour ward and informed consent obtained prior to enrollment into the study. In addition, participants’ case files were extracted and the quality of care the participants received audited, to assess the clinical management particularly data related to six standards of care (administration of intravenous fluids, intravenous antibiotics, monitoring of maternal vital signs, bladder catheterization, delivery within two hours, and blood grouping and cross matching). Interviewer-administered questionnaires were used to assess socio-demographic data, referral status and other relevant obstetric history related to the clinical management. Participants were interviewed for details where there was need for clarity in cases of under or over documentation of clinical findings. For each participant, the management received was compared to the recommended practice of the set standard of care. The study criteria (standard of care) for managing mothers with obstructed labour included; delivery by emergency caesarean delivery, destructive vaginal or assisted vaginal delivery within two hours of making a diagnosis of obstructed labour, intravenous access and at least one litre, Intravenous fluids to be given to correct metabolic derangement before delivery, Intravenous antibiotics to be administrated pre-operatively within one hour before any intervention to relieve the obstruction, blood grouping and cross-matching, monitoring temperature, fetal heart rate, pulse rate, blood pressure in an observation chart at least every four hours and bladder catheterization. To explore healthcare providers’ perceptions of the audit findings, ten semi-structured in-depth interviews (with one specialist obstetrician, one resident, the in-charge of the labour ward, the in-charge of theatre and the in-charge of the postnatal ward) and three focus group discussions (FGDs) (separately with midwives, intern doctors and residents) were conducted after each audit. Each FGD had ten participants and lasted about 40–80 min. The questions were thematically related to the criteria-based audit and were open ended. The interviews and FGDs were transcribed, coded and analyzed by thematic analysis. Using Kish Leslie formula (1965) for sample size estimation, a mean care score in the initial P1 and second audit P2 of 81.7 and 93.5 % respectively (from a study conducted in South-western Nigeria [4], an acceptable error margin M of 5 %, a power of 80 % and using the formula n = (Τ2[Ρ1(1 − Ρ1) + P2(1 − Ρ2)])/ Μ2 with as the standard value of 1.96, the minimum sample size of 324 participants for both audits was computed. Data entry was performed with EPI-DATA 3.1 and analyzed using STATA version 12. Results from initial and second audits were compared using chi square (χ2) for categorical variables and the Student t-test for numerical variables, and odds ratios computed. The level of statistical significance was set at p < 0.05. The performance score was computed as the number of participants that received the recommended divided by the total number of participants ×100.