Setting
This is a retrospective analysis of data from a prospective observational study conducted at HLH. HLH provides comprehensive emergency obstetric and basic newborn care 24 hours a day, seven days a week. The labour ward has six delivery rooms with one delivery bed each, and one operating theatre where CSs take place. Data were collected from all delivery rooms and the operating theatre.
Data collection and variables analysed
Trained research assistants have observed every delivery in the labour ward, working two-three in each shift covering 24 hours a day, seven days a week, using a structured data collection form. The research assistants were intensively trained in live observations and accurate data collection and reporting. Data collection for this study took place from February 2010 through January 2017, but data collected between 1.7.2013 (introduction of ambulance fee) and 01.01.2014 (additional introduction of delivery fees) were not included into the analysis. Data were collected prospectively during this study period, and there was a data quality control system to ensure the validity. Information collected included pregnancy complication, labour process and outcome, newborn information, and birth attendant information. For this study we only looked at factors describing perinatal characteristics and risk-factors not related to clinical management. In detail, the following variables were analysed: number of births per month, multiple births, antenatal care, gestational age, birth weight, fetal presentation (cephalic or non-cephalic), and macerated stillbirths. In addition, we included the following variables which are partly related to clinical management: spontaneous vaginal deliveries (SVD), fetal heart rate status on admission, and labour complication comprising obstructed labour, vacuum extraction, CS, pre-eclampsia/eclampsia, bleeding before birth, uterine rupture, and cord prolapse.
Analysis of gestational age revealed a large inter-rater variability before and after the introduction of delivery fees due to training of the healthcare workers in estimating gestational age in 2013, and the variable was thus discarded from the analyses.
Statistical analysis
Data was compared between the period before introduction of fees from February 2010 through June 2013 (41 months) and the period after introduction of fees from January 2014 through January 2017 (37 months).
Count data are presented as numbers and percentages and continuous data as means and standard deviations. Potential indicators for high-risk deliveries in the period before respectively after the introduction of fees were compared by either Chi-square test or Student’s t-test, as appropriate.
To detect and quantify potential changes in proportions of high-risk deliveries at HLH in the period after the introduction of fees, we constructed a Variable Life Adjusted Display (VLAD) plot (9), presenting the cumulative sum of observed numbers for each variable in the post-introduction period, minus the expected numbers if the situation without introduction of fees had persisted. The VLAD plot can then be interpreted as the cumulative excess or deficiency of risk factors over time, compared to the pre-introduction period. Variables included into the VLAD-plot were selected due to statistically significant differences in the period before versus after introduction of fees and to clinical relevance. The VLAD-plot comprised the following variables: total number of births, abnormal and not measured fetal heart rate status on admission, fetal presentation other than cephalic, SVD, labour complications, CSs, birth weight below 2500 grams, and birth weight above 4000 grams.
To verify statistical significance of the findings of the VLAD-plots and quantifying when enough information to claim a significant change had been accumulated, corresponding cumulative sum (CUSUM) plots were also constructed (10,11). These plots have a formal signal limit where the process is deemed to have demonstrated a persistent change when the signal limit is crossed. These CUSUM plots were constructed such that they were able to detect an increase in the proportion for those variables showing an increasing tendency and a decrease for those variables showing a reduction. The signal limits where chosen such that the processes would give a false alarm (type I error) at most once per 10 year on average.
Ethical considerations
This study was approved by the National Institute for Medical Research (NIMR), Ministry of Health in Tanzania (the HBB CQI program Ref. NIMR/HQ/R.8a/Vol.IX/1247 and the Safer Births project Ref. NIMR/HQ/R8a/Vol. IX/1434), and by the Regional Committee for Medical and Health Research Ethics, Western Norway (Ref. 2009/302 and Ref number 2013/110/REK). Data is not openly available, and we received permission by NIMR for this study and to publish the findings. All relevant health care providers were informed about the different HBB CQI and Safer Births quality assessment studies and gave oral consent. Patients were also informed about ongoing studies, but consent was not required for this descriptive study.