Indicators for measuring health promotion practice among healthcare workers in the Nelson Mandela Bay Municipality, South Africa: A cross sectional study

Background: Measuring indicators for health promotion (HP) practice among healthcare workers is essential if health goals and objectives must be achieved. Such indicators provide connections between health policies and health outcomes; and ultimately add value to healthcare. This study identied indicators of health promotion among healthcare workers and compared them across levels of healthcare facilities in Nelson Mandela Bay Municipality, South Africa. Method: A cross sectional study involving 495 healthcare workers randomly sampled from 23 hospitals including primary, secondary, and tertiary level hospitals was conducted. Questionnaires were distributed to medical doctors, nurses, and allied health workers (AHW) such as social workers, physiotherapists, occupational therapists, and speech therapist. Questions (hereto referred to as dimensions) in the questionnaires were categorized to address Facility Related Indicators (FRI), Health worker related indicators (HRI) and Outcome related indicators (ORI). Descriptive and bivariate analysis were used to identify the indicators of HP among the three HCW groups from the three levels of health care. Indicators observed to be signicant in the bivariate analysis were subjected to a multivariate analysis using the multinomial regression model. (p–value < 0.05). Results: Emerging Indicators were grouped into three categories: facility related indicators (FRI), healthcare workers’ related indicators (HRI), and outcome related indicators (ORI). Four FRI dimensions were observed to be predictors of HP among doctors. Two dimensions were positively associated with HP practices while two others were negatively associated with HP practices among medical doctors and AHWs. On the other hand, seven HRI dimensions were signicantly associated with HP among medical doctors and AHW. Furthermore, 5 outcome related indicator (ORI) dimensions were signicant predictors of HP among medical doctors while only two dimensions were predictors of HP among AHW. The generalized Hosmer–Lemeshow Chi-square test showed that the models for the different HP indicators t the data. Conclusions: We identied in Nelson South Africa. HP budget, HP communication channel, coordination, and policy and guidelines. Some of the emerging HRI dimensions included guidance on diet & lifestyle, adequate knowledge on patient condition, follow up and home visits; and ORI dimensions such as reductions in diseases and disabilities, health inequities and improved need assessment. Exploration of relationships between dimensions and various levels of health facilities revealed correlations between HCWs, policy and guidelines at the primary healthcare level. The results further showed association between HCWs and community-based placement (CBP)at the primary level; and HCWs with CBP, home visits and follow up at the secondary level.

.28% (n=21) 20.16% (n=26) 6.25% (n=1) 6.25% (n=1) 6.25% (n=1) 10.48% (n=35) 9.28% (n=31) 12 positively associated with HP among AHW. These factors were 0.75 and 1.8 times higher compared to nurses, respectively. On the other hand, HP Budget (Coeff: -0.726, 95% CI: -1.361 --0.091) and Provision to access patients HP need (Coeff: -1.798, 95% CI: -3.428 --0.167) were negatively associated with HP practices among AHW ( Table 2). The generalized Hosmer-Lemeshow test Chi-square was 16.13 with p-value 0.444, indicating a good t of the nal model. Ten dimensions comprising -education on disease condition, guidance on diet & lifestyle, need for routine check-up, adequate knowledge on patient condition, tness and health screening, assist with welfare services, participation in HP training, coordinated HP training for staff, home visit, community-based placements, and follow up post discharge were associated with HP practice among HCWs (See Additional le 2). At health facility level, eight dimensions were statistically signi cant indicators of HP at tertiary level, three at the secondary level, and only one at the primary level ( Table 3). The indicator dimensions that were statistically signi cant at the different healthcare levels are listed in Table 3.    Eight indicator dimensions focussing on outcome related indicator were identi ed. The dimensions that were statistically signi cant in the study area were reduction of number of diseases, reduction of number of disabilities, and reduced health inequities. Others are improved needs assessments and improved community participation and mobilization (See Additional le 3). Further analysis at the facility levels showed that three indicator dimensions; reduction of number of diseases, reduction of number of disabilities and improved need assessment were statistically signi cant at the tertiary healthcare level. However, no indicator dimensions were statistically signi cant at both primary and secondary healthcare levels ( Table 5).   affect the optimal functioning of HP within the health care system. Our results are in concordance with previous reports that have linked the adequate and established channels of communication from within the organization to enhanced health promotion. For instance, Fassl (28) showed that organizations and work environment thrive when leaders adopt effective channels of communication. This also suggests that the success of the organizations and work environment are dependent on the availability of effective leaders that are skilled and con dent to transform the vision into reality for improved public health.
Hao & Yazdanifard (29) also concluded that leadership and a clear organization are vital for successful operations (29). To enhance the realization of dimension that improve health promotions, there is need for increased coordination and effective channel of communication. According to Mohr et al (30), enhanced care coordination is largely attributed to favourable patient experiences, cost management and cost effectiveness. This further suggests that FRI has the potential to in uence other indicator categories.
Budget was another signi cant FRI dimension identi ed in our study. In South Africa, the public health system is funded through the general tax revenue to provide universal coverage to all citizens. Our results showed that participants emphasized the importance of nancial allocation through adequate HP budgets. However, most of the respondents reported absence or unawareness of any nancial budget for HP activities in the study area. Furthermore, the health system which is dependent on general tax revenue has also been affected by severe austerity measures leading to ineffective functioning of public hospitals. This also may affect the implementation of health promotion policies thus having long term implications on the disease prevention and health promotion outcomes. Financial budget is an all-important driver of health at all levels of care (31)(32)(33). This unfavourable response on HP budget may imply a possible lack of structural plan for HP at the policy level or lack of understanding of the potentially accruable bene ts from HP.
This study further highlighted low involvement in HP induction training by medical doctors and AHW. The observed low or absence of HP induction training reported by the HCWs may translate to poor HP knowledge by HCWs hence, inadequate HP practices. A recent study by Liu et al (34), which explored the experiences of health-care providers during the COVID-19 crisis in China emphasised the impact of training in enhancing the communication skills and preparedness of HCWs in healthcare management. Furthermore, previous studies reported signi cant improvement in con dence, knowledge and attitude of employers following a workplace health promotion training programme (35). Health care workers' poor participation in HP training or unawareness of its existence in their facilities, may suggest inadequate awareness of HP programmes in the study area. This compares well with ndings from a previous study which showed poor HP awareness among HCWs in the facilities (36). However, regarding educating patients on their disease condition, all HCWs fared well in this dimension. This may be because most information communicated to patient in this dimension focuses on biomedical evidence which forms part of the core professional training and not necessarily on HP. Most of the medical doctors and AHW reported absence or ignorance of HP guidelines and policy in their facilities. The observed the poor participation in HP practices among HCWS in the study area may be due to lack of or inadequate HP guidelines and policy. A recent study observed that poor knowledge of HP policy among most HCWs may contribute to poor HP practice among HCWs (37). This emphasizes the need for HCWs to have standardized guideline, policy, and programme appraisal to ensure e cient HP practices by HCWs at all levels.
Findings from the study further showed that HCWs responses regarding home visits to promote patients' health was poor. A previous study showed that most physicians indicated that home care should be provided to patients who are bedridden, incapacitated or patients lacking family support (38). According to these authors, the physicians who participated in the study showed dissatisfaction concerning abuse of services suggesting the need for de ned guidelines and policy to protect the interest of both patients and HCWs. Our ndings corroborate with a study done in Canada which showed a very good rate of patients' follow-up post discharge (39). However, the Canadian study reported a higher rate of follow-up of patients from bigger urban settings compared to those from the rural areas. The variations observed between home visits and follow-up in our study may suggest that most patients make conscious effort to come back to the hospital, and when they do, they must be attended to. The problem with this is that several patients may be unable to come back to the hospital due to cost, poor health, or other reasons and are therefore excluded from HP care and empowerment. This undermines the goal of Universal Health Care (UHC). For the agenda of UHC to be achieved, there must be a shift from the status quo that focuses only on patients visiting the health care facilities to HCWs visiting patients (40).
Furthermore, reduction in disease and disability were identi ed as HP outcome measures in the current study. This compares well with a previous study in which Maredza et al (41) showed that the best approach to reduce the burden of stroke is through HP. Reduction in health inequity also emerged as a dimension in measuring HP outcomes and this may be traceable to the historical apartheid background of South Africa that demarcated people along racial lines. The socioeconomic impact of this systemic divide is still evident in the South African healthcare system (42) as the quality of care differs along socioeconomic strata.
To our knowledge, this study is the rst to identify HP indicators among healthcare workers in the study area and this is the rst time a comparison of HP indicators among HCWs have been compared among different healthcare professionals and at the various levels of health care facilities.
Since this study focused on public hospitals, its ndings may have limitations to provide information regarding the private sector. Furthermore, the study was carried out in one municipality and hence the results may not be directly extrapolated to all other municipalities in the country. We therefore recommend further studies to explore the healthcare workers HP practices and perceptions using triangulation methods and possibly in other provinces in South Africa to providing further insight.

Conclusion
Healthcare systems worldwide are increasingly embracing health promotion as a strategy to attaining universal health coverage. Performance measurement of healthcare workers' health promotion practice using a set of indicators is central to achieving desired goals. Our study has not only identi ed these indicators but highlighted areas that need improvement. The relevance of individual dimensions classi ed as FRI, HRI and ORI in this study provides a template for the potentials in healthcare workers to be maximally harnessed. Adoption of these indicators will empower people to take control over their health while reducing cost. Since healthcare workers within health facilities have lasting impact on in uencing patients and their relatives' behaviour, this study strives to make a case for HP in healthcare facilities.
We however note that the pathway to institutionalizing measuring indicators may be a long one considering the rigours required of government to address manpower needs, deliberations, setting benchmarks, and de ning expected goals for HCWs. This study is therefore a contribution towards initiation of that change process.

Declarations
Ethics approval and consent to participate The study prior to data collected received ethical approvals from the University of KwaZulu Natal Biomedical Research Ethics committee (BREC), reference -BREC /00000088/2019, and the Eastern Cape Health Research committee, reference -EC_201910_012. Gatekeeper approvals were obtained from the respective Healthcare facilities. A written informed consent was received from all the participants before proceeding with the study.

Consent for publication
Not applicable.