An Outcome Evaluation of LifeNet Partograph Training in Masaka, Uganda

Background: Prolonged and obstructed labor is a leading cause of maternal mortality and morbidity in low resource settings. To prevent prolonged labor, the Uganda Ministry of Health (MOH) recommends using the partograph to monitor mothers in labor. Published literature has reported low rates of partograph use in Uganda, as well as improved partograph use after training. This study aimed to evaluate the effectiveness of LifeNet International’s (LN) partograph training in rural health clinics in Maska, Uganda. Additionally, the study sought to identify factors potentially related to partograph use, and thus to inform future implementations to increase partograph use in these low-resource settings.

Methods: LifeNet works with rural clinics in Uganda by providing training and management strategies to improve healthcare quality for mothers during delivery. In 2017, LifeNet began collaborating with the Duke Global Health Institute (DGHI) to evaluate LN’s impacts in six clinics in Masaka District, Uganda. As part of this evaluation project, this study is evaluating the impact of LN partograph training using direct observation, medical chart data, and facility-level data collected by LN. Additionally, semi-structured interviews were conducted by a DGHI researcher. The pre-training data were collected from May 15th to July 17th, 2017 and post-training from August 23rd, 2017 to January 29th, 2018 for this study. Follow-up direct observation data are scheduled to be collected from May 21st to July 26th, 2018. Quantitative data were analyzed using Stata version 14.2. Interview transcripts were reviewed for themes of health providers’ partograph knowledge and challenges of partograph use in practices.

Results: Before the LN partograph training, an estimated 19.8% of deliveries (42 of 212 observed) in study clinics were monitored with a partograph. A diagonal line drawn on the partograph helps the clinician to recognize possible labor complications (i.e. the action line). Sixteen (38.1%) of those that used partographs reached the action lines, among which five (31.2%) had actions under taken. In the first month after the LN partograph training, partograph use increased to 46.8% and was sustained for the remainder of the observation period. The proportion of partograph use did not change over time after the training (prevalence risk ratio, PRR=1.00, 95%CI: 1.00-1.00). Among all partographs reviewed after the training (n=594), health providers gave two interventions to manage abnormal labors. Mean duration of labor and proportion of prolonged labor did not change over time (risk ratio, RR=1.00, p = 0.561; RR=1.00, p=0.757, respectively). However, mean duration of labor was significantly higher among deliveries in which a partograph was used, compared to deliveries in which no partograph was used (RR=4.39, p<0.001). Furthermore, the proportion of deliveries with prolonged labor was higher in the partograph use group compared with the group that did not use the partograph, but the difference was not statistically significant (RR=5.97, p=0.072).

Based on the interviews with clinical providers in these clinics, there seems to be some education in use of the partograph in their schooling; however, there remained some misunderstanding about partograph use and interpretation. Health providers indicated that lack of accessibility to blank partographs in clinics, heavy workload, and lack of periotic check were challenges in using partographs to monitor labor.

Conclusions: Partograph use increased following the LN training and was sustained for at least five months afterwards. This type of clinical training program may be effective in improving maternal healthcare quality in Last Mile health facilities in resource-poor settings, like Masaka, Uganda.