Does public service motivation matter in Moroccan public hospitals? A multiple embedded case study

Background: The motivation of health workers is a key concern of policy makers, practitioners and researchers. Public Service Motivation (PSM), defined as the altruistic desire to serve the common interest, to serve others and to help patients and their families regardless of financial or external rewards, has been shown to be key to the performance of public servants. Yet, limited attention has been paid to this kind of motivation in health care settings in low- A nd middle-income countries. Little is known about PSM and its contextual specificity in the Moroccan health system. We set out to qualitatively explore the meaning of PSM and its expression among health workers in four public hospitals. Methods: We adopted a multiple embedded case study design to explore PSM in two well-performing and two poor-performing hospitals. We carried out 68 individual interviews, eight focus group discussions and 11 group discussions with different cadres (doctors, administrators and nurses). We carried out thematic analysis using NVivo 10. Results: Our analysis shows that public service motivation is a notion that seems natural to the health workers we interviewed. Daily interactions with patients catalysed health providers’ affective motives (compassion and self-sacrifice), a central element of PSM. It also provided them with job satisfaction aligned with their intrinsic motivation. Managers and administrative personnel express other PSM components: Attraction to public policy making and commitment to public values. A striking result is that health workers expressed strong religious beliefs about expected rewards from God when properly serving patients. Conclusion: This study highlights the presence of PSM as a driver of motivation among health workers in four Moroccon hospitals, and the prominence of intrinsic motivation and compassion in the motivation of frontline health workers. Religious beliefs were found to shape the expression of PSM in Morocco. This study is part of a larger research project that adopted the realist evaluation approach [53] to examine the links between leadership, staff motivation and performance. This study is exploratory in nature [54] in a sense that it explores the motivation of health workers using insights from PSM theory in the context of Morocco. Morocco is a lower middle-income country with a population of 35.6 millions [55]. Islam is the religion of most Moroccans and this has an impact on daily life and work place practices [56–58]. The Moroccan society is multi-cultural, collectivistic and strongly attached to family relationships and filial piety [56, 57]. In health, there has been a significant progress in many health indicators (e.g. 35% reduction of the maternal mortality rate between 2010 and 2016), the extension of the coverage to the poor and vulnerable populations, the decentralisation and the introduction of public private partnerships [59, 60]. However, the Moroccan health system remains weak, ranked 134th out 195 countries in terms of health access and quality of care [61] with an inequitable access to care and a poorly regulated private sector [62, 63]. The Moroccan health system is constrained by an acute health workforce shortage, for instance having 0.7 doctors per 1.000 inhabitants and 0.92 nurses and midwives per 1.000 inhabitants) [64, 65]. Studies indicated a lack of staff commitment, poor motivation, low job satisfaction and poor working conditions [58, 66] that have hampered the implementation of many health system reforms, for instance quality assurance programmes [67, 68], fee exemption policies [69]. Other studies found that these health workforce issues affected the quality of patient-provider interactions [70, 71]. In this study, we adopted the case study design. The case here is PSM as experienced by health personnel in hospitals. The multiple embedded case study design is appropriate to the exploration of complex phenomena, such as PSM in real world settings and allows for comparison between sites [72]. We purposefully selected two high-performing and two low-performing hospitals. We used the results of the national quality assurance programme called concours qualité to select hospitals as our study sites. This programme assessed hospital performance in eight dimensions: (1) accessibility/availability/continuity; (2) patient security and responsiveness; (3) satisfaction; (4) ethics; (5) quality assurance; (6) resource utilisation; (7) technical competencies and (8) leadership. The overall performance score index is measured by the ratio between the actual score assessed by the external audit and the maximum obtainable score. We refer to [67, 68] for an overview of the programme. Using data from the quality assurance report 2011 [73] and 2016 [74], we identified hospitals with a significant increase or decrease of performance between 2011 and 2016. In the field of public management, the definition of PSM has evolved since 1990, broadening from ‘individual predisposition’ to a more detailed description (Table 1). Definitions of PSM In the last two decades, most PSM research focused on developing measurement scales. These studies were carried out in the USA, Europe, Asia and South America [78]. Wright noted a high degree of variability in operational definitions and a diversity of PSM scales [79]. To overcome this diversity, Kim and colleagues [80] refined the multidimensional scale developed by Perry and Wise [35] and validated it across 12 industrialised countries [81–83]. It comprises: In this study, we consider individual motivation as mix of motives including a continuum that ranges from extrinsic motivation to more autonomous forms of motivation (intrinsic motivation and PSM). We mean by intrinsic motivation ‘the doing of an activity for its inherent satisfaction rather than for some separable consequences’ [84]. We adopt Ryan and Deci’s definition of extrinsic motivation: “a construct that pertains whenever an activity is done in order to attain some separable outcome” (e.g. tangible and verbal rewards) [84]. We adopted the definition of Vandenabeele of PSM. As explained below, we used the 4 components of PSM as defined by Kim et al. [83] and the concepts of intrinsic and extrinsic motivation as defined by Ryan and Deci [84] to analyse the data. We started with interviews and then conducted group discussions. A document review was carried out all along the study. All data were collected during the period January–June 2018 (see Tables ​Tables22 and ​and33). Data collection Respondents characteristics We started the data collection by interviewing hospital staff. In total, we carried out 68 in-depth interviews (IDI) with senior-, middle-, operational-level managers and health professionals [85, 86]. We explored the views of respondents on public service motivation, its definition and its expression, as well as the factors that may influence PSM. We used adapted open-ended interview guides for each category of respondents (Additional file 1). These were tested in a pilot study with professionals. We carried out the interviews until saturation was attained. In a second stage, we carried out eight focus group discussions and 11 group discussions with different cadres (administrators, nurses and doctors) to further explore the key constructs mentioned by interviewees in relation to motivation. This allowed us to deepen the analysis across different categories of health workers (Table 3). The FGD and group discussions were conducted by the first author, who used a facilitator guide (Additional file 2). We conducted the FGD following standardised procedures described by Krueger and Morgan and used probes, asked follow up questions, summarised key themes and sought verification from participants [87, 88]. We used qualitative purposive sampling [86] in order to identify respondents for the in-depth interviews and the focus group discussions. Interviews were carried out in Moroccan dialect. All interviews, FGDs and group discussions were audio-recorded with the exception of two interviews. In these specific cases, we took notes and transcribed the unrecorded interviews using memory recall [85]. All transcripts were checked for accuracy by two co-authors (ZB and BM). Table 2 presents the breakdown per hospital. We used codes to identify respondents anonymously, referring to the hospital EJMH, NHMH, RKMH, SMBA. At the end of each contact with research participants, we wrote a brief contact summary that included major themes and ideas arising from the interaction following guidance provided by Miles and Huberman, and Krueger [87, 89]. Table 3 represents the summary of data collection tools and respondents’ profiles (Additional files 3, 4, 5 and 6 provide detailed descriptions of the respondent characteristics). In order to identify key elements in the broader health policy context and to describe the general context, we collected documents all along the study. Key informants at the four hospitals and at the Ministry of Health contributed to identifiying relevant documents. We focused on human resources availability and skill mix data, strategic plans, audit documents and quality assurance reports. We structured the qualitative data analysis along the analytic phases of compiling data, interpreting data, discussion and drawing conclusions [90]. During the initial coding cycle, we coded all data sources (transcripts, contact summaries and field notes) using different coding techniques (concept, hypothesis and “in vivo” coding) [91] and used NVivo 10 v11. 4.3 software to manage the data [92]. During the second coding cycle or pattern coding, [89], we used the four PSM components (attraction to public service, commitment to public values, compassion, self-sacrifice) described by Kim and colleagues [83] and the intrinsic/extrinsic types of motivation defined by Ryan and Deci [84]. As we will present below, we also identified other categories of motives that emerged from the inductive analysis of data (religious based motives). The coding was discussed during research team meetings. These meetings were conducted at different moments during the analysis, focusing on the initial coding, the second coding cycle, the in-case analysis and the cross-case analysis. The research protocol was approved by the Moroccan Institutional Review Board in Rabat (n°90/16) and the Institutional Review Board of the Institute of Tropical Medicine, Antwerp (N° 1204/17). All interviewees were informed before the start of data collection about the study objectives, the topics, the type of questions and their right to refuse being interviewed or interrupt the interview at any time. The same information was included in an information sheet that was given to candidate interviewees and reiterated when the written consent form was discussed before the start of the interview. The informed consent forms were signed by the participants and co-signed by the researcher. A copy of the signed consent form was given to research participants.