Study design and population
A cross-sectional pre-training survey using a modified Attitudes Toward Health Care Teams Scale (ATHCTS)[9] was conducted among health care providers undergoing a two-day interactive interprofessional HIV training in Kumasi and Agogo from November 2019 to January 2020. The training comprised eight HIV specific modules reflecting commonly encountered clinical or programmatic challenges developed by a panel of leading health educators from across AFREhealth network and the University of California, San Francisco [1]. The main training was preceded by training of trainers to facilitate the subsequent training. Trainees were from the following health professions; Medicine and Dentistry, Pharmacy, Physician Assistantship, Nursing and Midwifery, and Medical Laboratory Sciences. Health care workers were selected from one teaching and five district hospitals in the Ashanti Region: Komfo Anokye Teaching Hospital (KATH), Maternal and Child Health Hospital and Suntreso Government Hospital, in Kumasi; Bekwai Municipal Hospital; Obuasi Government Hospital and Agogo Presbyterian Hospital. The health care professionals comprised new and old providers. New providers were newly qualified health care professionals within 12 months of graduation. Hence the new providers included house officers, interns, and rotation nurses/midwives. Old providers were professionals who had been working for more than 12 months post qualification. The training was conducted in English; all trainees were proficient in English.
Study sites
KATH is the second-largest hospital in Ghana, and one of the main tertiary referral health facilities in the northern sector of the country. It is a 1,200-bed capacity hospital with 12 clinical directorates including the Directorate of Medicine. The hospital trains various categories of health professionals including nurses and midwives, medical and dental students, pharmacy students, laboratory scientists, house officers, and postgraduate resident doctors. The Department of Medicine has several sub-specialties including the Infectious Disease Unit which has a specialized HIV clinic. Patients at the HIV clinic are managed by a team of physicians, nurses, pharmacists, and other health care providers. HIV clinics were held twice and three times a week for adults and children respectively. An average of 300 and 80 new adult and childhood cases of HIV were attended to at the clinic each year. Other directorates such as Obstetrics and Gynaecology also attend to cases in conjunction with physicians from the unit. The other hospitals were the first referral hospitals (district hospitals) in their respective districts/metropolitan area. All the hospitals had specific clinics for people living with HIV with each facility/clinic attending to an average of 1000–2500 cases (about 100–300 new cases) a year. Except for the Agogo Presbyterian Hospital, all the other hospitals were being supported by the (United States) President’s Emergency Fund for AIDS Relief (PEPFAR) at the time.
Study procedures
We engaged key stakeholders in the partner institutions and facilities from the onset and throughout the planning and implementation of the programme. All health professionals, house officers and interns of the five health professions in KATH, and health care professionals involved in HIV care in the selected hospitals were eligible for the training. Within KATH, the number of participants from each profession was selected by the programme/unit head in proportion to their number in the unit. For health care workers in the selected hospitals, 12 professionals working in HIV care were selected by their unit heads for the training.
To assess trainees’ attitudes towards interprofessional health care teams, participants were asked to complete an online Google form (Google, Inc., Mountain View, CA, USA) pre-training survey in English using a modified ATHCTS with 14 items by Curran et al. [9]. This modified scale was adapted from Heinemann et al. [10], who identified three main factors as influencing attitudes namely quality of care, cost of team care and physician centrality, comprising 14, seven and six items respectively [10]. For the modified ATHCTS, Curran et al. selected 11 items from the quality of care factor and three items from the costs of team care factor as appropriate for pre-licensure students with little or no experience with items relating to physician centrality [9]. We chose the 14-item modified scale since that is recommended for assessing the attitudes towards health care teams among a wide variety of health professions [9, 11, 12]. The survey was typically completed within the 24 h preceding the training. The survey tool included sections on trainee characteristics (age, gender, profession, health facility, and professional experience) and the 14 items on the adapted ATHCTS. Responses to the 14 items were scored on a five-point Likert scale ranging from one (strongly disagree) to five (strongly agree). Three items regarding time constraints which are worded such that agreement represents negative attitudes were reverse scored. Total scores ranged from 14 to 70 with higher scores indicating more positive attitudes toward interprofessional health care teams.
Statistical analysis
Data was summarised using descriptive statistics of mean and standard deviation for continuous variables, and frequencies and percentages for categorical variables. An exploratory factor analysis was conducted to categorise the 14 items of the modified ATHCTS. The suitability of the data for factor analysis was assessed. The correlation matrix showed several coefficients were ≥ 0.3 indicating high correlation among items for factor analysis [13]. The Bartlett test of sphericity (p < 0.001) and Kaiser–Meyer–Olkin measure of sampling adequacy (0.79) confirmed strong correlation for application of dimensionality reduction [13, 14] among the 14 items in the modified ATHCTS, and the adequacy of the sample for factor analysis respectively. Factor extraction was performed using principal component analysis and factors with eigenvalues > 1 (Kaiser’s criterion) were retained. Factor rotation was performed using Varimax rotation and items with factor loadings of at least 0.4 were considered to contribute to that factor [14, 15]. Internal consistency was assessed using Cronbach’s alpha with a threshold of 0.7 [16]. The overall Cronbach’s alpha for the 14 modified items was 0.71. The overall mean score was estimated by adding all 14 items; all negatively worded statements were reverse scored. Shapiro–Wilk normality test was used to check the normality of the data; the overall attitude score was not normally distributed (p = 0.005). Hence, the Wilcoxon rank-sum (Mann–Whitney) and Kruskal–Wallis tests were used to test the mean attitude difference among the demographic characteristics. All statistical analyses were performed using Stata 17.0 (StataCorp, Texas, USA), and p < 0.05 was considered statistically significant.
