Lived experiences of frontline healthcare providers offering maternal and newborn services amidst the novel corona virus disease 19 pandemic in Uganda: A qualitative study

Background The COVID-19 pandemic has brought many health systems in low resource settings to their knees. The pandemic has had crippling effects on the already strained health systems in provision of maternal and newborn healthcare. With the travel restrictions, social distancing associated with the containment of theCOVID-19 pandemic, healthcare providers could be faced with challenges of accessing their work stations, and risked burnout as they offered maternal and newborn services. This study sought to understand the experiences and perceptions of healthcare providers at the frontline during the first phase of the lockdown as they offered maternal and newborn health care services in both public and private health facilities in Uganda with the aim of streamlining patient care in face of the current COVID-19 pandemic and in future disasters. Methods Between June 2020 and December 2020, 25 in-depth interviews were conducted among healthcare providers of different cadres in eight Public, Private-Not-for Profit and Private Health facilities in Kampala, Uganda. The interview guide primarily explored the lived experiences of healthcare providers as they offered maternal and newborn healthcare services during the COVID-19 pandemic. All of the in depth interviews were audio recorded and transcribed verbatim. Themes and subthemes were identified using both inductive thematic and phenomenological approaches. Results The content analysis of the in depth interviews revealed that the facilitators of maternal and newborn care service delivery among the healthcare providers during the COVID-19 pandemic included; salary bonuses, the passion to serve their patients, availability of accommodation during the pandemic, transportation to and from the health facilities by the health facilities, teamwork, fear of losing their jobs and fear of litigation if something went wrong with the mothers or their babies. The barriers to their service delivery included; lack of transport means to access their work stations, fear of contracting COVID-19 and transmitting it to their family members, salary cuts, loss of jobs especially in the private health facilities, closure of the non-essential services to combat high patient numbers, inadequate supply of Personal Protective equipment (PPE), being put in isolation or quarantine for two weeks which meant no earning, brutality from the security personnel during curfew hours and burnout from long hours of work and high patient turnovers. Conclusion The COVID-19 Pandemic has led to a decline in quality of maternal and newborn service delivery by the healthcare providers as evidenced by shorter consultation time and failure to keep appointments to attend to patients. Challenges with transport, fears of losing jobs and fear of contracting COVID-19 with the limited access to personal protective equipment affected majority of the participants. The healthcare providers in Uganda despite the limitations imposed by the COVID-19 pandemic are driven by the inherent passion to serve their patients. Availability of accommodation and transport at the health facilities, provision of PPE, bonuses and inter professional teamwork are critical motivators that needed to be tapped to drive teams during the current and future pandemics. We conducted this embedded qualitative study as part of a bigger study that assessed the impact of COVID-19 pandemic on the provision of Maternal and Newborn healthcare services in eight health facilities in Kampala, Uganda between June 2020 and December 2020 [32] during the first phase of the lockdown. We used the phenomenological [33, 34] and inductive thematic approaches [35] to explore the lived experiences and perspectives of 25 healthcare providers as they offered maternal and newborn services in the eight selected facilities in Kampala using in depth interviews. This study was conducted in eight health facilities (two Private hospitals, three Private-Not-for Profit hospitals and three Public health facilities) in Kampala, Uganda. These eight facilities were purposively selected because they are the biggest service providers in the three sectors (public, private-not-for profit and private) offering maternal and newborn health care in Kampala. All of the eight health facilities had most of the different cadres of healthcare providers for maternal and newborn health with brief description provided in Table 1. Prior to participant recruitment, we sought permission from the different hospital institutional review boards. After obtaining permission, we met the different hospital administrators who later allowed us to meet the healthcare providers in maternal and newborn health based on their availability and convenience. We purposed to meet healthcare providers of different cadres offering maternal and newborn health services. These included obstetricians/gynecologists, theatre in-charges, nurse midwives, medical doctors, ward in-charges, nurse in-charges of immunization, antenatal, postnatal and family planning clinics. The selected healthcare providers were then given two contacts of the Principal investigator and the research team. We purposively interviewed 25 healthcare providers at the eight selected health facilities using in depth interviews which were preferred to focus group discussions to minimize any spread of the pandemic. Disinfection protocols were observed prior to the interviews. All the in depth interviews were administered in English, the official language used in Uganda in quiet rooms at the different selected health facilities as recommended by the hospital administrators. Healthcare providers actively involved in maternal and newborn health service delivery at any of the eight selected health facilities during the study period that consented to participate in the study were recruited. Healthcare providers involved in maternal and newborn health services at the eight selected health facilities who were on leave or inaccessible physically during the study period were excluded. We had three teams on the study. Team 1 was in charge of data collection. The team was composed of two researchers and two field note takers. The two researchers had doctoral degrees and were familiar with the local hospital settings. This team had research training for three days. They were trained on how to identify and interview potential participants. They were also trained on participant recruitment while observing the research ethics in accordance to the Declaration of Helsinki [36]. The two field note takers were fluent in English and Luganda, the locally spoken language. Team 2 was in charge of data analysis. It was composed of Principal investigator and one administrator. This team had to ensure transcription accuracy and data analysis. Team 3 was composed of two independent researchers whose task was checking rigor according to the Lincoln—Guba criteria [37]. After obtaining informed consent from the participants, 25 healthcare providers had in depth interviews by two doctoral degree level interviewers between June 2020 and December 2020. All interviews were administered in English. Two note takers captured the participants’ non-verbal expressions with their consent in addition to the field notes. After ascertaining data saturation with no new emerging themes, we stopped the data collection [38, 39]. The interviews lasted between 45 to 90 minutes. The interviews captured the participant socio-demographic information, the way they perceived service delivery before and during the COVID-19 pandemic, the facilitators and barriers to quality maternal and newborn service delivery and their recommendations to optimal service delivery in future disasters or pandemics using open ended questions. Whenever clarity was needed, more specific questions were raised by the interviewers so that all the required information was collected. All of the interviews were tape recorded. Two interviewers and two note takers were trained prior to the data collection. A pilot study was carried out with four healthcare providers to pretest and modify the interview guide. Data from the pilot study was also included in the analysis as the healthcare providers in the pilot were not included in the main study. The interviews were tape recorded and transcribed verbatim immediately after the interviews. The transcriptions were compared with field notes throughout the study period. We ensured that the coordinators of the interviews or discussions didn’t participate in the analysis but critiqued the results from the analysis and ensured that these results conformed to their expectations from the discussions. This was done to validate the study findings and also ensure quality in the study. Field notes and transcripts, codes and their interpretations were made by separate teams of investigators. Data was backed up on hard drives, online databases and two computers. The research materials were kept under restricted access by only authorized staff for participant confidentiality and privacy. We obtained ethical approvals from The AIDS Support Organisation (TASO) Institutional Ethics review board, (TASOREC/064/2020-UG-REC-009), and Uganda National Council of Science and Technology (HS924ES). We also obtained administrative clearances from the eight health facilities. Verbal and written informed consents were obtained from all study participants after an elaborate explanation of the study. Participants were reimbursed for participating in the study in form of transport refunds. Participants were reassured that participating in the study was voluntary and that they could opt out of the study without compromising the relationship with the research team. Confidentiality and participants rights were observed throughout the study. All participants’ data (audiotapes, records, transcripts and notes) were kept in a secure location accessible only to study personnel. Study participants were identified by pseudonyms rather than actual names in the final report. Using the inductive thematic approach [35] and Colaizzi’s process of data analysis for phenomenological studies [34, 39], the research team took the following steps: data was prepared by typing out the interviews, thereafter using sentences, phrases or paragraphs, generated meaning units from the context of the participants’ voices. We then converted the concepts generated into codes (text coding) using semantic tags. The primary codes were then generated and meaning units shortened to formulate ‘compressed meaning units’. We later revised the text codes comparing similarities and differences between the codes thereby integrating what appeared as similar codes. We then critically looked at all the transcript steps and codes and classified them based on their relationships or differences. We ensured reliability of the codes, and then revised the classes. Data was coded and analyzed manually using a framework matrix developed using an Excel workbook built after a detailed and careful process of the emerging codes. We kept comparing the codes from the data generated. Similar codes were put into subcategories and these subcategories were later put into the main themes. In cases of disagreements, the research team had to discuss until an agreement was reached. To ensure rigor in the data collection, we used Guba and Lincoln criteria [37], that included, data credibility, confirmability, transferability and dependability. Triangulation was checked by team 3 that was devoted to continuous reading through of the transcripts to ensure ongoing comparison of the key information generated from one hospital to another during the data collection and analysis processes. Dependability was observed by the stringent coding procedure and inter-coder corroboration. We made sure to document what each code meant in detail as illustrated in Table 3. Data confirmability was observed by ensuring that participants’ statements were captured with barely any modifications made. Data transferability was ensured by the research team so that a rich, thick description of the study process was documented to enable replicability in a similar context elsewhere [40]