The impact of strike action by Ghana registered nurses and midwives on the access to and utilization of healthcare services

Background As the largest professional group, nurses and midwives play instrumental roles in healthcare delivery, supporting the smooth administration and operation of the health system. Consequently, the withdrawal of nursing and midwifery services via strike action has direct and indirect detrimental effects on access to healthcare. Objective The current study examined the impact of strike action by nurses and midwives with respect to access to and use of health services. Method Data were collected retrospectively from a total of 181 health facilities from all the 16 administrative regions of Ghana, with the support of field officers. Because the strike lasted for 3 days, the data collection span three consecutive days before the strike, three days of the strike and three consecutive days after the strike. Data analysis was focused comparing the utilization of healthcare services before, during and after strike. Data were analysed and presented on the various healthcare services. This was done separately for the health facility type and the 16 administrative regions. Findings The results showed that; (1) the average number of patients or clients who accessed healthcare services reduced drastically during the strike period, compared with before the strike. Majority of the regions recorded more than 70% decrease in service use during the strike period; (2) the average number of patients or clients who accessed healthcare services after the strike increased by more than 100% across majority of the regions. Conclusion The study showed that strike action by nurses and midwives negatively affected access to and utilization of healthcare services. Data were gathered from health facilities located across the 16 regions in Ghana and across all the levels of healthcare. As noted previously, the healthcare system in Ghana is structured. At the highest level is the teaching hospitals, followed by regional hospitals, district hospital, polyclinic, health centers or clinic and CHPS. Data were collected from health facilities that are under the auspices of Ministry of Health and Ghana Health Service (Collectively referred to as Government facilities) and Christian Health Association of Ghana (collectively termed CHAG facilities). Private health facilities were excluded from the study since the working conditions of nurses at these facilities differ from the counterparts in Government or CHAG facilities. In this study, majority of the facilities were owned by government (n = 155, 85.6%), whereas 26 (14.4% CHAG facilities. A total of 112 field officers were recruited across the 16 regions in Ghana through the regional offices of the GRNMA. A dedicated WhatsApp platform was created for the field officers and the research team to facilitate communication relating to the project. A training workshop was organized and held via the Zoom videoconference platform at the convenience of the field officers who were working as nurses in their respective health facilities. The major area for the training was the data gathering process, including how to maintain data integrity, avoid data contamination as well as ensure ethically responsible research conduct. This was intended to ensure quality data gathering and transmission via a dedicated electronic portal powered by Google. The study variables, operationalized as the services rendered by the health facilities, were decided by the research team members following a series of meetings and consultations with researchers, policy makers and practitioner. The team also took into consideration local health priorities and the demands of the Sustainable Development Goal (SDG) 3. The research team unanimously agreed on the following study variables; (1) outpatient department services, (2) admissions, (3) deliveries, (4) surgical services, (5) reproductive health services, and (6) antenatal clinic (ANC) services. We defined ANC services as healthcare services delivered to pregnant women. While these may include reproductive health services, we also note instances where reproductive health services are delivered to non-pregnant women. Therefore, in this study, we focus on reproductive health services as services accessible to non-pregnant women. The inclusion of delivery services, for instance, was in accordance with indicator 3.1 of the SGD3 (reducing “global maternal mortality ratio to less than 70 per 100,000 live births”) and indicator 3.2 (reducing neonatal mortality to at least as low 12 per 1,000 live births……”). Reproductive health service was also included in view of indicator 3.7 of the SDG3: “universal access to sexual and reproductive health care services, including family planning……”. Lastly, the focus on antenatal services reflect indicator 3.1 of the SGD3 which is to reduce “global maternal mortality ratio to less than 70 per 100,000 live births” and indicator 3.2 that is concerned about reducing neonatal mortality to at least as low 12 per 1,000 live births and under-5 mortality to at least as low as 25 per 1,000 live births.” Data were collected before, during and after the strike period to allow for comparative analyses and discussions across different data collection points, with reference to the strike period. Because the strike lasted for three consecutive days (21st to 23rd September, 2020; Monday to Wednesday), we collected data for the 3-day period to appreciate the scale of the impact of the strike action. Besides, the number of clients or patients who accessed healthcare services differ from day to day. This made it difficult to restrict the data collection to any of the days the strike action occurred. To obtain a baseline data against which to assess the impact of the strike, we collected data for the 3 consecutive days before the strike action (14th to 16th September 2020; Monday to Wednesday). Lastly, data were collected 3 consecutive days after the strike action was called off (28th to 30th September 2020; Monday to Wednesday) to estimate the use of health services following the suspension of the strike action. The data collection lasted for approximately 2 months, spanning 10th October to 3rd December 2020. Data collection was aided by the tool designed by the research team based on the pre-determined study variables discussed previously. Prior to the data collection, institutional permission was sought from the various health facilities with introductory or permission letter issued by the GRNMA national secretariat to the field officers. As stated previously, data were collected for three consecutive days for each data collection period (e.g., before, during and after strike). The field officers completed the hardcopy of the questionnaire for each facility. Thereafter, they were provided with a dedicated link, powered by Google Form, where they inputted and transmitted the data electronically to a centralized receiver accessible to the research team. The same questionnaire was used across the health facilities. The field officers were informed to input nil or zero where the data sought for does not exist. For example, because CHPS compounds do not conduct surgeries, data on this service will not be available. The electronic form requires that the field officers provide additional information on the region, district, type of facility (e.g., hospital or health center) and ownership of facilities (e.g., Government or CHAG) where data were collected. Regular updates were provided on the WhatsApp page to keep the field officers informed about the submissions received. This is a retrospective study in which data on the number of people who utilized various healthcare services before, during and after strike action by nurses and midwives were gathered. The study did not involve direct human subject engagement. Rather, data were obtained from institutional archives as an aggregate data. The focus was on how many people visited or utilized healthcare services within the time frame above, without focusing on the background or demographics of users of healthcare services. The data collected were also devoid of identifying information relating to the facilities. This means that neither the facility nor clients/patients will be identified. Thus, data were fully anonymized. The data on number of people accessing healthcare services is notably a public data in Ghana. The project was underpinned by other relevant ethical considerations in research, including confidentiality, data safety and data protection. Access to data was restricted to the research team or other individuals supporting the project, mainly data analysts. These individuals signed a statement confirming that they would adhere to the study procedures regarding confidentiality. The data collected were analyzed as regional aggregate data to further delink the healthcare facilities. The Institutional Review Committee of the Research and Grant Institute of Ghana has declared that given the nature and type of data collected, ethics approval prior to data collection was not necessary. By the end of the data collection process, a total of 191 submissions were received. However, some of the submissions were duplicates, perhaps because of the technical and internet connectivity issues. It was also observed that, some field officers did not provide the exact or absolute number of service users for the study variable. Instead, they provided inscription such as “over 400”, making it difficult to determine the exact number of service users under reference. The dataset was subsequently cleaned by deleting the data anomalies or deviations, leaving a total of 181 submissions for analyses. The analysis of data proceeded on two key assumptions; (1) health facilities under the various categorization (e.g., hospital, health centers) in a region will be similar with respect to the average number of service users than those outside the region. That is, hospitals in Ashanti region will be similar in terms of the number of service users than hospitals in Volta region. This assumption is centered heavily on the variations in the population distribution across the regions which in turn influence the number of health service consumers; and (2) health facilities falling under a particular category will be similar in terms of the range of the services provided. For example, it was assumed that CHPS compound across the country will offer virtually the same type of health services. In the same vein, hospitals across the country are more likely to render the same set of health services. Any difference should be subtle or negligible. Based on the foregoing, data was analyzed at the health facilities level, segregated by region. That is, hospital data were analyzed on regional basis as were data from health centers. To proceed, we computed the average number of service users for each region, taking into consideration the type of health facilities. For example, data from the hospitals in Ashanti region were summed and divided by the total number of hospitals that provided data for the study. This resulted in the average number of health service users from hospitals in Ashanti region. In instances where data were available for only one type of health facility in a region, the same data were used since the mean could not be calculated. Although the mean is sensitive to outliers, it is the most widely used descriptive statistics in research and publication. To address problems relating to outliers, we aggregated and analyzed data along regional framework and by nature of health facilities. Data was prepared using the IBM SPSS Version 23 and analyzed using excel. The analyses involve mostly descriptive statistics. We computed the percentage change in the average number of individuals accessing health services before, during and after strike.