Background
HIV/AIDS has an impact on health sector workers in various ways. It increases fear of infection at work, changes or increases tasks and workload, and increases the emotional burden and stress levels of health workers [
1‐
3]. However, little concrete evidence exists of the impact of HIV/AIDS on the health sector [
4], as few studies have been conducted.
The Ugandan health sector is the main provider of HIV/AIDS-related services for a population of about 29 million people. The HIV prevalence rate is 6.7 per cent among adults (15–49 years) and about 900,000 adults and 110,000 children (0–14 years) live with HIV [
5]. About 88 per cent of the Ugandan population live in rural areas [
6]. The health sector faces staff shortages. A total of 30,000 health workers were employed in 2004, and yet an extra 5000 qualified staff were required [
7]. Most staff are nursing assistants, a cadre with minimal professional health qualifications [
7]. A shortage of health workers could negatively influence access to and quality of care. Therefore, retention and motivation of remaining staff is crucial.
Retention of health workers is linked to job satisfaction, which is influenced by various factors such as physical working conditions, relationships with colleagues and managers, pay, promotion, job security, and responsibility, although priorities will differ in different contexts [
8]. Job satisfaction influences motivation to work but is not sufficient on its own. When someone is satisfied with his or her job, he/she is not necessarily motivated to perform well. Motivation is defined as 'an individual's degree of willingness to exert and maintain an effort towards organizational goals ([
9]: p.1255-6).' Factors that influence motivation to perform well need to be identified in each context. They may include opportunities for promotion or training, opportunities for an increase in remuneration, receiving appreciation for work or obtaining recognition from managers, colleagues and patients.
Published studies about the impact of HIV/AIDS on the health sector and its workers in low-income countries focus mostly on occupational hazards [
10,
11] and on knowledge, attitudes and practice [
12‐
21]. Some studies focus on a variety of aspects such as stress and burnout, working conditions, knowledge and attitudes, and organizational support [
22‐
26]. Although the studies vary in design and are, therefore, difficult to compare, they confirm that health workers fear infection, face stress and burnout and are concerned about being stigmatized. The main causes identified include a lack of knowledge and skills and a lack of organizational support to deal with fear, stigma, stress and burnout, and changes in tasks and responsibilities.
Studies in Uganda [
27‐
30] have shown that HIV/AIDS influences patient care and that it increases health workers' fear of infection for various reasons. However, a comprehensive overview of the perceptions of managers and health workers of the impact of HIV/AIDS is missing. This is required when designing country-level strategies. Identifying the influence of HIV/AIDS on staff motivation – to design activities that mitigate the impact of HIV/AIDS and integrating these activities in strategies for motivation and retention – is crucial to assure quality of care.
At the time of the study Uganda had 56 districts, served by government, private for-profit and private not-for-profit (PNFP) health facilities. There were 10 regional referral hospitals and 45 district hospitals run by the government. PNFP facilities accounted for 43 per cent of the hospitals and 24 per cent of the lower health care facilities, mostly in rural areas. Many of the PNFP facilities provided health services and trained health workers, and 78 per cent of these facilities were faith based [
31].
This article describes the results of a study exploring the effects of HIV/AIDS on hospital staff and organizational responses to mitigate these effects in four different general hospitals in rural districts of the Central Region in Uganda. The study aimed 'to identify the influence of HIV/AIDS on staff working in general hospitals at district level in rural areas and to explore support required and offered to deal with HIV/AIDS at the workplace'.
Methods
Research design and research questions
The study design was exploratory and cross-sectional and consisted of a quantitative and a qualitative component. It intended to answer the following research question:
▪ What are the perceptions of hospital managers and staff regarding the effects of HIV/AIDS at their workplace?
▪ What are the current overall working conditions in the selected hospitals?
▪ What are the current support measures offered and required in the selected hospitals to assist staff in dealing with HIV/AIDS at work?
▪ Are hospital staff motivated to work, and to what extent does HIV/AIDS influence staff motivation?
Theoretical framework
No standard theoretical framework for the impact of HIV/AIDS on hospital staff exists in the literature, therefore the research team developed its own framework. Our main hypothesis was that staff would be discouraged from working because of low motivation and stress, which are influenced by HIV/AIDS and by general working conditions, and that this would contribute to low performance of health systems. According to Chen et al (2004), enhancing the performance of health systems requires, in addition to adequate financial and material resources, workforce objectives on coverage, competence and motivation [
32]. Staff motivation is, therefore, an important contributing factor to the performance of health systems. Different theories exist about motivation. In this article health workers' motivation is explained using expectancy theory, adapted to the workplace by Vroom and by Porter and Lawler [
33]. This theory describes staff motivation as a rational process of decision making. It explains that staff will be motivated to work if they believe they can perform a task successfully when they put effort into it, if they believe that the outcome will be positive when they perform the task, and if this outcome is valued by them [
34]. Our study focused on the impact of HIV/AIDS on motivation. It explored whether staff perceived effects of HIV/AIDS on their work, what these effects were and whether organizational efforts were in place to mitigate this impact, and it identified factors motivating staff to work. Subsequently, during analysis, we tried to identify a relationship between staff motivation and the perceived impact of HIV/AIDS. We also tried to identify if there were differences in perception and responses between the staff of the four hospitals.
Sampling and study population
Four general hospitals in rural districts were selected purposively, as these were the facilities that provided most HIV/AIDS services close to the population. Purposive sampling is a sampling method used in qualitative research to select 'a limited number of informants strategically so that their in-depth information will give optimal insight into an issue about which little is known ([
35]: p.199).' The hospitals were located in four different rural districts in the Central Region in Uganda, a region with a 9.4 per cent HIV prevalence rate, one of the highest in the country [
36]. This region consists of four urban and 12 rural districts. In these rural districts comprehensive HIV/AIDS services are provided by 13 public and nine faith-based, private not-for-profit hospitals [oral communication from MoH]. The hospitals were selected according to their type, as different types of hospitals have different working conditions and a different working environment, which might have an impact on staff perceptions and experiences. We were also interested in exploring whether or not individual hospital settings influenced experiences and perceptions, even though hospitals had the same background, so we intended to include more than one hospital of the same type in the sample. Time and budget allowed us to conduct the study in four hospitals: two public and two faith-based. In each health facility, the study population consisted of all health workers, support staff which came in direct contact with patients or patient fluids, and managers.
For the survey, the quantitative component of our study, quota sampling took place, recruiting an appropriate number (quota) in each category of health staff, maintaining a proportional representation of health staff in the sample. Quota sampling means that a quota is set for each attribute (in this case the type of professional cadre), and the quotas are set 'so that they represent together the known distribution of the control attributes across the known population ([
37]: p.37).' We intended to include as many health workers as possible, aiming to interview between 30 and 50 per cent of the health workers employed in the selected hospitals. Among support staff cleaners were selected, because they work on the wards and come into contact with patients and with patients' fluids. As they were not the key respondents in the research, we only interviewed a limited number of cleaners. In total, 237 members of staff were recruited according to their availability (presence and having time to be interviewed), from a total of 594 staff employed by the hospitals at the time of the study and in direct contact with patients or patients' fluids. Table
1 shows the distribution of different staff categories interviewed.
Table 1
Composition of the survey sample
Support staff | | 36 (15%) | 10 (13%) | 8 (17%) | 12 (17%) | 6 (14%) |
Clinical staff | Allied health professional | 27 (11%) | 6 (8%) | 5 (11%) | 8 (11%) | 8 (18%) |
| Enrolled Nurse/Midwife | 74 (31%) | 23 (30%) | 14 (30%) | 19 (27%) | 18 (41%) |
| Doctor | 8 (3%) | 3 (4%) | 2 (4%) | 1 (1%) | 2 (5%) |
| Nursing assistant | 70 (30%) | 28 (36%) | 11 (24%) | 24 (34%) | 7 (16%) |
| Registered Nurse/Midwife | 22 (9%) | 7 (9%) | 6 (13%) | 6 (9%) | 3 (7%) |
| Sample of clinical staff as percentage of total number of clinical staff in hospital | | 65%* | 46% | 39% | 35% |
Respondents for the qualitative component were recruited purposively, using maximum variation sampling. In each hospital, four managers, six health workers from different departments and one or two support staff were selected. Health workers were selected for interviews from among those with experience in working with HIV/AIDS patients, with implementing HIV/AIDS-related tasks and general caring tasks. If a person was not available, either a new appointment was made or another person with similar tasks was selected. Support staff were interviewed according to their availability. In total, 44 respondents were interviewed, and 25 respondents participated in four focus group discussions.
Data collection
Quantitative data were collected by research assistants who had experience in conducting interviews and who were trained on the background of the study and the research questions. Data were collected through interviews, using a structured questionnaire. Questions were asked about infection control guidelines, availability and use of protective materials, treatment and disposals of sharps, perceived risk of HIV infection at work, occurrence of injuries, and support offered and used to prevent and deal with HIV infection. Additionally, questions were asked about working conditions, supportive supervision, training in HIV/AIDS tasks, and staff motivation. Data were collected during a period of two weeks in September 2005.
Qualitative data were collected by experienced researchers through in-depth interviews and focus group discussions, using a topic guide. Open questions were asked about the impact of HIV/AIDS on work, dealing with HIV/AIDS at work, and support required and offered by the hospital. Questions were also asked about general working conditions and staff motivation.
Data analysis
Quantitative data were analysed with Epi-info 6.1 and Stata 9.2, using non-parametric tests (Chi-square and Kruskal-Wallis test) for descriptive analyses. Assuming that the members of staff working in a particular hospital were not completely independent, Generalized Estimated Equation (GEE) models were used to determine independent factors associated with not being motivated. Motivation was dichotomized: 'not being motivated' included the categories: indifferent, discouraged and very discouraged, and 'motivated' included the categories motivated and very motivated to the question 'How motivated are you in your work?' Factors associated with not being motivated in univariate analysis (P < 0.10) were selected for multivariate analysis. The final model was created using stepwise backward selection of variables and was checked for confounding. Only the variables that showed a significant effect or acted as confounder were kept in the model. Key variables for analysis included fear of infection, injuries and actions taken, protective materials in place and used, support services to prevent and deal with HIV infection, and general working conditions. Qualitative data were analysed using a framework developed according to the research objectives, key issues and themes. Issues for analysis included the impact of HIV/AIDS at work, dealing with difficulties because of HIV/AIDS (such as fear of infection, stigma, emotions and workload), support required and offered at work, general working conditions, and staff motivation.
Ethical considerations and quality assurance
The quality of data was safeguarded by using experienced interviewers, assuring the confidentiality and privacy of respondents, and by asking informed consent and permission to tape interviews and focus group discussions. All research instruments were pre-tested. The research team members who conducted the in-depth interviews and focus group discussions were involved in the development of the research protocol and data collection instruments. They also participated in data analysis and report writing. Research assistants were trained to use the structured questionnaire and interviewed under supervision of an experienced researcher.
The validity of data was assured by triangulation. Data were triangulated by:
▪ asking the same questions during focus group discussions and in-depth interviews;
▪ exploring the same topics among support staff, health workers and managers; and
▪ comparing and contrasting information from the interviews and focus group discussions with information from the questionnaires and with the registers and reports from the hospitals.
The protocol was approved by the Ethics Review Committee of Uganda Martyrs University.
Discussion
In this study staff and managers reported that HIV/AIDS has an impact on workload, leads to changes in tasks and affects emotions, although the latter was less pronounced in this study. Injuries were reported to be common, and most staff feared infection at work. Respondents knew colleagues who were HIV-positive, although HIV-positive staff remained in hiding, and staff did not want to get tested due to fear of being stigmatized. No HIV-positive staff talked openly about their HIV status. The reported impact of HIV/AIDS demonstrated in these Ugandan hospitals corroborates published studies elsewhere: frequent occurrence of injuries, reported by 36 per cent of respondents in our study, is reported by 57 per cent of respondents at the central hospital in Uganda [
29], and ranged between 26 and 53 per cent in studies reporting on injuries elsewhere [
22,
24,
26]. Fear of infection, in our study reported by 83 per cent of respondents, varied in two Ugandan studies between 30 and 47 per cent [
27,
28], and our study corroborates reported fear of infection from studies elsewhere [
12,
22,
24,
26]. In our study 77 per cent of those respondents who felt adequately protected feared getting infected. These findings corroborate two studies among doctors in Nigeria, which show that feelings of fear of infection persist, despite the availability and use of protective materials [
38,
39]. A number of studies in other countries with a high HIV prevalence rate demonstrated that staff felt stressed and faced burnout, often being emotionally exhausted [
14,
22‐
26]. This was less pronounced in Uganda. Studies in Zambia [
22,
26] reported, as in our study, that HIV-positive staff are not willing to tell others about their status and that health workers in general are unwilling to be tested.
In the Ugandan hospitals, organizational responses to the impact of HIV/AIDS were implemented haphazardly. None of the hospitals had written policies to prevent and mitigate the impact of HIV/AIDS and to support HIV-positive staff. Organizational responses were reported to consist of informing staff about infection control, making protective materials and existing HIV/AIDS-related services available, although in none of the hospitals respondents reported that these services were clearly communicated to staff. Areas that were not explicitly addressed in any of the hospitals were stigma, HIV counselling and testing among staff, supporting HIV-positive staff, availability and use of PEP, and emotional support. Lack of organizational support is also shown in studies in Zambia [
22,
26]. The findings show that management needs to urgently address the impact of HIV/AIDS in health facilities. Workplace HIV/AIDS policies need to be designed and implemented, and use could be made of the generic guidelines developed by ILO/WHO [
40] and of workplace policies that have been designed for the private sector in Uganda [
41].
We explored the relationship between the perceived impact of HIV/AIDS and staff motivation. Motivation appeared to be determined, among others, by working conditions, such as overtime, frequency of supervision, provision of HIV counselling, and use of ARVs by staff. These factors are greatly influenced by HIV/AIDS. It might be that because of a lack of strategies to support HIV-positive staff, staff using ARVs do not feel motivated. This could not be explored, as no HIV-positive staff came forward during our study.
Staff with a higher level of education in particular were less often motivated than other staff, which might be linked to the reported lack of qualified staff and, therefore, having more responsibilities. Men are less often motivated than women in the survey, but it is not clear why, and answers from the interviews did not confirm this.
The most important reported reason for staff motivation was 'liking the work', and salaries and financial benefits appeared less important. This is corroborated elsewhere, and various studies [
42‐
47] show that, although financial incentives are important, other motivating factors were appreciation, recognition and career possibilities. A number of reported reasons for motivation, such as 'liking the work', 'recognition', 'teamwork', and 'salaries and financial benefits' were not included in the multivariate analysis, as no separate questions were asked with respect to these variables. These could have been determinants or confounders for staff motivation and would need to be included in further studies.
As low motivation of health workers contributes to poor health worker performance and thus affects quality of care, Human Resources Management (HRM) activities to improve staff motivation need to be implemented. Managers should be aware that there is a complicated relationship between motivation and performance. According to expectancy theory, motivation to perform is a combination of feeling able to successfully perform a task when putting effort into it, obtaining a positive outcome (reward) upon completion, and valuing this outcome. This means that health facility managers need to implement HRM activities and use leadership skills to:
▪ assure that the expected level of performance is discussed and agreed upon by staff and management;
▪ support staff in such a way that they feel able to achieve the expected level of performance;
▪ assure that expected positive outcomes of performance (eg financial or non-financial rewards) outweigh expected negative outcomes (e.g. being tired and overworked); and
▪ assure that expected rewards are provided when performance is achieved [
48].
A combination of interventions in all these areas is likely to lead to motivation for performance.
Our study identified that staff and managers considered HIV/AIDS to be constraining their work, as it either led to a perceived negative outcome (such as fear of getting infected while delivering care) or had an impact on their perceived ability to provide quality care (due to increased workload, emotional stress, changes in tasks and limited training in new tasks). Integrating activities to prevent and mitigate the impact of HIV/AIDS into existing HRM activities, instead of developing a 'vertical' HIV/AIDS workplace programme, can improve these perceptions. Examples of this type of activity are the integration of discussions on infection control, stigma and difficulties with HIV/AIDS patients into staff meetings and daily supervision; including HIV/AIDS-related topics in educational sessions to improve staff knowledge and skills; and including support to HIV-positive staff in workplace policies for chronically ill staff. Workload issues can be addressed by improving teamwork, rotating tasks, and taking measures aimed at staff attraction and retention. Our study identified motivating and discouraging factors among staff, but ranking these factors is required to assist managers to prioritize and align incentives for performance with valued positive outcomes of staff performance.
Caution has to be taken to replicate strategies without adapting these to the prevailing context. Although the type of hospital (public or PNFP) did not influence the reported impact of HIV/AIDS or organizational responses, the specific hospital context did seem to influence the perceptions and experiences of hospital staff, although differences were not always statistically significant. Two examples to illustrate this (Table
5 and
6):
In hospital B, staff were generally less motivated. The reported working conditions were less positive than in the other hospitals: hospital B had the highest bed:staff ratio, the lowest availability of five types of protective materials, the lowest number of staff that received daily supervision, irregularly organized educational sessions and the highest number of staff reporting not receiving compensation for overtime. In addition, it was one of the two hospitals with the highest number of injuries, and respondents of hospital B had the lowest knowledge of PEP and its availability in the hospital. Lastly, although pay was higher in hospital B than in hospitals C and D, staff motivation in hospital B was lower than hospitals C and D. A focus for interventions would be to analyse leadership and management and to improve upon available activities such as supervision and educational sessions.
On the other hand, in hospital A many respondents were motivated despite the fact that the most overtime was reported in this hospital and a low number of staff received compensation. Hospital A had a high number of respondents being aware of guidelines, knowing about and using HIV/AIDS-related services for staff and being aware of PEP offered in the hospital. In this hospital a high number of respondents reported being supervised daily. However, in hospital A staff complained about a heavy workload due to extra tasks related to HIV/AIDS. This is in contradiction to the bed:staff ratio and the reported number of new patients in the outpatient department and might be linked to reported extra tasks related to HIV/AIDS such as RCT. The focus for intervention might be on evaluating staff experiences with RCT so as to better adapt these to staff capabilities.
It is important to know that although general measures should be designed to mitigate the impact of HIV/AIDS and to motivate staff, these examples demonstrate that each hospital management team needs to have a different focus in strategies. They show that differences occur in leadership and management, availability of resources and organization of services between hospitals working in similar conditions. Management needs skills and support to analyse the working conditions in their facilities and to adapt generic guidelines to their own specific situation.
Informing policymakers which interventions are successful under which circumstances and for which staff groups is important, as it allows the formulation and implementation of evidence-based approaches [
49]. Various authors [
50,
51] describe the importance of taking the context and process into consideration when formulating and implementing interventions to address performance problems. Identifying strategies to address factors contributing to performance problems, such as low motivation, is important but managers need to be aware that blue print solutions do not exist. Health systems are social systems which are open and thus are influenced by and interact with their context. Additionally, the way interventions are implemented depends on the vision, skills and experiences of stakeholders involved in its implementation (management, health workers, support staff, district teams etc). This is also the case for strategies for staff motivation in the health sector. Therefore, evidence building needs to include information on the process of implementation, the context and any changes in the context. Randomized trials, which for health system interventions are considered by many the most credible designs for evidence building [
50], normally do not include data collection on context and process. To answer the question 'what works for whom and why', these trials need to be complemented with different types of data and of data collection methods, such as methods to describe practitioners' views on lessons learned and conditions for success. Up to now little has been written on what works and what does not with respect to staff performance, their motivation and retention in the health sector in low-income countries (among others, [
52,
53]). Experiences with activities to mitigate the impact of HIV/AIDS and the integration of such activities into motivation and retention strategies for health workers are remarkably scarce in literature. There is an urgent need to document and share experiences with interventions to motivate and retain staff in low-income countries and with activities mitigating the impact of HIV/AIDS on the health sector and its workers.
Study limitations
We were not able to measure stress, as no validated instrument for Uganda exists. Therefore, the influence of HIV/AIDS on stress could not be identified. Additionally, for some members of the research team it was difficult to probe on sensitive topics such as personal experiences with HIV/AIDS, HIV status of respondents and colleagues, and emotional feelings, which might have influenced data collection. The survey was based on the availability of staff, which could have caused a selection bias and which we were not able to check as data on absence and sick leave were not available in the hospitals. However, management assured us that absence and sick leave were not major problems faced in any of the hospitals and that staff were systematically scheduled to work in different shifts, without differences in profile. In addition, our own impression during the study was that absence due to stress, burnout or low motivation was not an issue in any of the hospitals. The questionnaire and the interview guide could have better addressed motivation by including questions in line with expectancy theory. Lastly, the results reflect staff opinion on their knowledge, skills and practice. Confirming if reported knowledge, skills and practice correspond with actual knowledge, skills and practice was not possible, due to time and budget constraints.
Authors' contributions
MD and VB were principal investigators of the study and were responsible for protocol development, study implementation, data analysis, and report writing. MD drafted the manuscript. EM, GN and OJ were members of the research team and contributed to protocol development, study implementation, data analysis, and report writing. VB, EM, GN and OJ commented on the first manuscript. MB analysed data of the quantitative study component, contributed to report writing and commented on the draft manuscripts. GvdW discussed the conceptual framework for the study and substantially commented on the draft manuscripts. All authors read and approved the final manuscript.