Are healthcare professionals ready for national health insurance? Knowledge, attitudes, and implementation-related perceptions at a Public Hospital in Gauteng, South Africa

¹Virginia Apandju Osako Ngende, Department of Health, Ekurhuleni Health District: Pharmaceutical Services, Germiston, South Africa, ²Africa Centre for HIV/AIDS Management, Faculty of Economic and Management Sciences, Stellenbosch University, Stellenbosch, South Africa

^&Corresponding author
Virginia Apandju Osako Ngende, Department of Health, Ekurhuleni Health District: Pharmaceutical Services, Germiston, South Africa

Introduction    

South Africa is embarking on one of the most ambitious health system reforms in its democratic history. Through National Health Insurance (NHI), the country aims to reduce longstanding inequities between the public and private health sectors and ensure equitable access to affordable, quality healthcare for all citizens, regardless of socio-economic status [1,2].

South Africa's healthcare system continues to show marked disparities in access, resources, and quality of care between the public and private sectors. Approximately 84% of the population relies on public healthcare services, despite having considerably fewer financial and human resources compared with the private sector [3,4]. These structural inequalities remain a major obstacle to achieving universal health coverage (UHC) [4,5]. Against this backdrop, the NHI represents South Africa's principal strategy for reforming the health system [1].

While the policy direction is clear, NHI readiness remains considerably more complex. South Africa is currently in the pre-implementation phase of the NHI, during which governance structures, regulations and implementation mechanisms are being developed [1,6]. Alongside these developments, persistent concerns remain regarding governance and accountability, financing, infrastructure, workforce shortages and organisational capacity [7,8]. Mofolo and Heunis [9] caution that, unless broader systemic constraints are addressed, the NHI is at risk of remaining an aspirational reform rather than evolving into a fully operational and sustainable healthcare reform.

Notwithstanding these overarching concerns, the NHI Act has been signed into law in 2024 [2]. Rollout is imminent. As it stands, the success of NHI and the realization of UHC will soon fall squarely on the shoulders of healthcare professionals - who will need to translate policy into everyday clinical practice. Healthcare professionals have long been recognised as the cornerstone of effective health systems. Their preparedness to embrace NHI policy reform is likely to influence implementation, service delivery and public confidence in the health system [8,10].

As the frontline implementers of the NHI, healthcare professionals will be expected not only to deliver healthcare services under a new financing and service-delivery model, but also to communicate with patients, adapt to new organisational processes and contribute to system-wide changes. In essence, understanding healthcare professionals' knowledge and attitudes towards the NHI, as well as their perceptions of the health system's readiness and their own preparedness for implementation, is therefore vital.

Previous NHI-related South African studies have provided valuable insights into awareness of the NHI, broad attitudes towards the reform and perceptions about health system readiness [8,10]. For example, both Mukwena et al. [10] and Mudzweda et al. [8] question the country's readiness for implementation. Mukwena et al. [10] identified organisational barriers - including inadequate infrastructure, workforce shortages, insufficient resources, and limited staff preparedness - as key constraints to implementation. Similarly, Mudzweda et al. [8] found that healthcare practitioners perceived hospitals as insufficiently prepared, citing staff shortages, inadequate infrastructure, limited training, poor communication, and resource constraints as the main barriers to a successful NHI rollout. To the best of our knowledge, however, existing research has, thus far, predominantly examined these NHI-related aspects independently, with limited attention to their interrelationships. Consequently, relatively little is known about how healthcare professionals simultaneously evaluate multiple aspects of the NHI, and how they relate to each other. Addressing this gap is important because healthcare professionals' understanding and evaluation of the reform may ultimately influence how the NHI is implemented in practice.

Hence, the aim of this study was to examine healthcare professionals' knowledge of the NHI, attitudes towards the policy reform, and perceptions of its anticipated effectiveness, organisational readiness, and implementation risks. Specifically, the study sought to examine the relationships among these implementation-related constructs, determine whether they differed according to professional and demographic characteristics, and identify factors independently associated with attitudes towards the NHI. In so doing, this study - conducted at a public hospital in Gauteng Province - aimed to generate baseline information to inform the development of targeted communication and stakeholder engagement strategies towards ensuring district-level NHI readiness. Moreover, the findings may provide a basis for future studies exploring potential differences between hospital- and clinic-based healthcare professionals, thereby generating more insights into their respective perspectives and readiness for NHI rollout.

Methods     

Study design and setting: a cross-sectional study was conducted at a selected public hospital in the eastern region of Gauteng province, South Africa. Data were collected via a self-administered paper-based and online survey consisting of closed-ended questions. The study specifically targeted healthcare professionals - nurses across all professional categories, medical doctors, and allied health professionals (which included pharmacy personnel, physiotherapists, dietitians, occupational therapists, radiographers, social workers, clinical psychologists, speech therapists, and audiologists).

Sampling: all healthcare professionals at the hospital were targeted for inclusion in the study. Data collection took place over four months. Potential participants were initially informed about the study during multidisciplinary meetings. WhatsApp groups were utilized as a vehicle to disseminate a questionnaire link and a QR scan code - which was created using Microsoft Forms - to facilitate and streamline participation. Hard copies of the questionnaire were also made available.

using G*Power, we aimed for a medium effect size, α = 0.05, β = 0.20 (80% power) [11,12]. The calculated sample size was considered in relation to the total population (N = 485) and the expected distribution of participants across the relevant categorical variables. Out of a target population of 485 healthcare professionals, 214 responses were received, yielding a response rate of 44.1%. Among these respondents, 203 (94.9%) consented to participate and were included in the final analysis, while 11 (5.1%) declined participation. The resulting final sample size (n = 203) was considered adequate for the planned analyses after assessment of the relevant statistical assumptions. The analyses included repeated-measures analysis of variance (ANOVA), Pearson correlations, one-way multivariate analyses of variance (MANOVAs), and multiple regression.

Data collection tool: the researcher-developed survey items were based on the 2023 National Department of Health's publication entitled "Understanding the National Health Insurance (NHI) - frequently asked questions" [13]. To enhance content validity and minimise potential measurement bias, the instrument was independently reviewed by two experts in the field of Public Health. The experts assessed whether each item appropriately reflected the construct under investigation and adequately responded to the research question. Each question's relevance was evaluated using a 4-point scale (1= not relevant; 2=somewhat relevant; 3=quite relevant; and 4= highly relevant). All questions were either rated as 3 or 4 and considered 'relevant'. Reliability was measured via internal consistency of the various combined scales and domains (see below) using reliability analysis (Cronbach's alpha).

The selection of study constructs/variables was informed by implementation science. The Consolidated Framework for Implementation Research (CFIR) identifies healthcare professionals' knowledge and beliefs about an intervention, together with organisational readiness and implementation climate, as important determinants of implementation success [14,15]. Accordingly, knowledge reflects understanding of the reform, attitudes indicate willingness to support it, perceptions of effectiveness and implementation risks capture judgements about likely benefits and barriers, while organisational readiness reflects confidence in the health system's capacity to implement change. Collectively, these constructs provide a theoretically informed assessment of factors likely to influence successful NHI implementation.

Measures

Demographic information: participants were asked to indicate their gender, profession, age group, years of work experience, and prior exposure to NHI (Table 1). As Table 1 shows, a total of 203 healthcare professionals participated in the study. Most respondents were female (89.2%) and worked as nurses (69.0%). Approximately one-third (33.5%) were aged 40-49 years, while the largest proportion (32.5%) had 0-5 years of work experience. Most participants (70.0%) reported low prior exposure to information about the NHI.

Knowledge score: knowledge was measured with 10 statements that could be either true or false. It covered the following themes related to the NHI: Core principles and objectives, funding and implementation, and common misconceptions. Each correct answer was scored as 1, while incorrect answers and no answers provided were scored as 0. The total score for correct answers could vary between 0 and 10.

Attitude toward NHI: attitude toward NHI was assessed using 8 items rated on a 5-point Likert scale (1 = strongly disagree to 5 = strongly agree). The items assessed overall support for NHI, healthcare equity, motivation to contribute to implementation, professional autonomy, facility preparation, willingness to recommend NHI, government-stakeholder collaboration, and financial protection for patients. The items were combined into a single attitude score, with higher scores indicating more favourable attitudes (α = 0.926).

Perceptions toward national health insurance: perceptions were assessed using 5-point Likert items (1 = strongly disagree to 5 = strongly agree) and grouped into 3 conceptually related domains: effectiveness and feasibility (4 items; α = 0.820), readiness and capacity (3 items; α = 0.795), and risks and concerns (3 items; α = 0.715). Higher scores indicated more favourable perceptions of NHI; negatively framed risk and concern items were reverse-scored before calculating the composite score.

Data analyses: data were exported from Excel and analysed using SPSS 30. Descriptive statistics included frequencies and percentages for categorical variables and means and standard deviations for continuous and composite variables (knowledge score, attitudes, effectiveness/feasibility, readiness/capacity, and risks/concerns). Pearson correlations examined associations between knowledge score, attitude, and the three perception variables. A repeated-measures ANOVA, with Greenhouse-Geisser correction and Bonferroni-adjusted pairwise comparisons, compared mean attitude and perception scores. A series of one-way MANOVAs, followed by univariate ANOVAs and Bonferroni-adjusted comparisons where appropriate, examined differences in knowledge score, attitude, and perceptions according to profession, age group, prior NHI exposure, and years of work experience. Finally, hierarchical multiple regression was used to identify factors independently associated with attitude towards NHI, controlling for relevant demographic, professional, exposure, and knowledge variables. Statistical significance was set at p < 0.05.

Ethical considerations: ethical clearance was obtained from Stellenbosch University's Research Ethics Committee: Social, Behavioural and Education Research (REC: SBE) (project number: 34853) and the Ekurhuleni Health District Research Committee (research project number: 06/10/2025/01). The study was also registered with the National Health Research Database (NHRD), with an NHRD number of GP_202510_002. An official permission letter to conduct the study in this hospital was granted by the hospital management team. Participants provided written or online consent (ticking a box) before the survey.

Results     

Descriptive findings descriptive statistics were used to summarise the five main study variables: knowledge score, attitude, and the implementation perception variables effectiveness/feasibility, readiness/capacity, and risks/concerns (Table 2). Participants generally demonstrated good knowledge of the NHI (M = 6.66, SD = 1.73). At the item level, knowledge score was strongest for its broader principles: 81.4% correctly identified universal health coverage as a core aim and 80.9% recognised its role in promoting greater equity between public and private healthcare. Knowledge was less consistent for operational aspects, including how services would be purchased (55.9% correct) and how the NHI would reduce out-of-pocket expenditure (57.4%).

Overall, attitude towards the NHI was moderately favourable (M = 3.31, SD = 0.87), and participants were similarly positive about its potential effectiveness and feasibility (M = 3.23, SD = 0.81). Perceptions of readiness and capacity were less favourable (M = 2.96, SD = 0.91). At the readiness/capacity item level, only 20.6% agreed that the hospital workforce was adequate to provide services under NHI, while 52.2% disagreed. For risks/concerns at the item level, the findings were particularly pronounced: 82.8% expressed concern about governance and accountability, 79.8% about payment or supply-chain disruptions, and 79.8% about increased workload without additional resources. Correspondingly, the risks/concerns (reverse-scored scale) had the lowest mean score (M = 1.85, SD = 0.71), indicating substantial concern about NHI implementation. Attitude and perception domains were scored from 1 to 5. For the reverse-scored risks and concerns domain, lower scores indicate greater concern.

Comparison of attitude and perceptions: next, a repeated-measures ANOVA with a Greenhouse-Geisser correction was conducted to determine whether participants differed in their ratings across four within-participant domains: attitude, effectiveness/feasibility, readiness/capacity, and risks/concerns.

There was a significant overall effect of domain, F(2.10, 424.27) = 288.20, p < 0.001, partial η² = 0.588, indicating that participants rated the four NHI domains differently. Bonferroni-adjusted comparisons showed that attitude and effectiveness/feasibility did not differ significantly (p = 0.239), whereas all other pairwise comparisons were significant (all ps < 0.001). Overall, participants were similarly positive about the NHI and its potential effectiveness, but less confident about implementation readiness and expressed substantially greater concern about implementation risks.

Relationships between knowledge score, attitude and perceptions: Pearson correlations were conducted next to examine relationships among the five main study variables: knowledge score, attitude, and the three implementation-related perception variables. As shown in Table 3, most of the study variables were significantly correlated. The relationships among attitude, effectiveness/feasibility, readiness/capacity, and the reverse-scored risks/concerns domain were positive, indicating that a more favourable attitude tended to accompany more positive perceptions of NHI effectiveness and implementation readiness, as well as fewer implementation concerns. The knowledge score showed a somewhat different pattern. This variable was positively related to attitude and effectiveness/feasibility, negatively related to the reverse-scored risks/concerns domain, and was not significantly related to readiness/capacity. Overall, a favourable attitude was closely linked to confidence in NHI's effectiveness and implementation readiness, while greater knowledge was accompanied by both a more favourable attitude and greater implementation concerns.

Differences according to participant characteristics: the next step was to conduct a series of one-way MANOVAs examining whether participants differed in their NHI-related knowledge score, attitude and perceptions according to selected demographic and professional characteristics. The five dependent variables were knowledge score, attitude, effectiveness/feasibility, readiness/capacity, and risks/concerns. Separate MANOVAs were conducted for four independent variables: profession, age group, years of experience, and prior exposure to NHI information. Where a significant multivariate effect was found, follow-up univariate ANOVAs and Bonferroni-adjusted pairwise comparisons were used to identify where the differences occurred.

Profession: there was a significant overall effect of profession on the combined dependent variables, Wilks' Ʌ = 0.847, F(10, 392) = 3.39, p < 0.001, partial η²= 0.079. Follow-up ANOVAs showed significant professional differences in knowledge score (F(2, 200) = 9.43, p < 0.001, partial η² = 0.086), and readiness/capacity (F(2, 200) = 4.37, p =0.014, partial η² = 0.042). Bonferroni comparisons showed that nurses had lower knowledge scores than doctors (p = 0.001) and allied health professionals (p = 0.009), while doctors reported lower readiness/capacity than nurses (p=0.011). No significant professional differences were found for the other variables.

Age group: age group also had a significant overall effect on the combined outcomes, Wilks' Ʌ = 0.808, F(15, 538.71) = 2.89, p < 0.001, partial η² = 0.069. Participants aged 20-29 years had significantly higher knowledge scores than those aged 40-49 years (p = 0.007). In contrast, participants aged ≥50 years reported a more favourable attitude than those aged 20-29 years (p = 0.021) and 30-39 years (p = 0.003). Participants aged ≥50 years also reported greater readiness/capacity than those aged 20-29 and 30-39 years (both p = 0.038).

Years of experience: years of experience did not have a significant overall multivariate effect on the combined outcomes, Wilks' Λ = 0.856, F(20, 644.38) = 1.55, p = 0.060, partial η² = 0.038.

Prior exposure to NHI information: prior exposure to NHI information had a significant overall multivariate effect, Wilks' Λ = 0.891, F(5, 197) = 4.83, p < .001, partial η²= 0.109. This difference was attributable to knowledge score: participants with high prior exposure had higher knowledge scores (M = 7.38) than those with low exposure (M = 6.35), F(1, 201) = 16.13, p < .001, partial η² = 0.074. Prior exposure was not significantly associated with the other variables.

Factors independently associated with attitude towards NHI: lastly, hierarchical multiple regression was conducted to identify factors independently associated with attitude towards NHI (Table 4). Attitude towards NHI was the dependent variable. Profession, age group, years of experience, and prior exposure to NHI information were entered in Step 1, followed by knowledge score in Step 2 (Table 3). Perception variables were not included in the regression model because of their strong correlations and conceptual overlap with the attitude outcome (which could introduce multicollinearity and obscure the independent contribution of the predictors).

Entering the demographic and professional variables in Step 1 explained 13.6% of the variance in attitudes, F(10, 192) = 3.04, p = 0.001, R²= 0.136, adjusted R² = 0.092. Adding the knowledge score in step 2 significantly improved the model, ΔR² = 0.119, ΔF(1, 191) = 30.65, p < 0.001. The final model was significant, F(11, 191) = 5.97, p < 0.001, R² = 0.256, adjusted R² = 0.213, explaining 25.6% of the variance in attitude.

Knowledge score was therefore the single strongest predictor of attitude in the final regression model (β=.388), contributing an additional 11.9% of explained variance beyond the demographic and professional characteristics. Participants aged 40-49 years and ≥50 years also reported a more favourable attitude than those aged 20-29 years. After adjustment for age group and the other predictors, participants with more than five years of experience reported a less favourable attitude than those with 0-5 years of experience. Profession and prior exposure to NHI information were not independently associated with attitude.

Discussion     

South Africa's National Health Insurance (NHI) aims to reduce healthcare inequities; however, its success will depend partly on the healthcare professionals expected to translate policy into practice. Previous South African studies have questioned implementation readiness, highlighting challenges related to staffing, infrastructure, resources, training and communication [8,10,16]. This study examined healthcare professionals' knowledge, attitudes, perception of NHI effectiveness, readiness and concerns regarding its implementation. The principal finding was that healthcare professionals could be broadly supportive of NHI and its potential benefits while simultaneously expressing considerable concern about the health system's readiness for its implementation. Greater knowledge was associated with more favourable attitudes, but also with greater awareness of potential implementation risks.

The descriptive findings provide the starting point for understanding this pattern. Participants generally understood the broad principles of NHI and held moderately favourable attitudes, but knowledge was weaker on some operational aspects. They appeared more familiar with what NHI is intended to achieve than with how it will operate in practice. This adds a knowledge dimension to earlier findings: Mukwena et al. [10] identified limited staff preparedness among factors constraining implementation, while Mudzweda et al. [8] highlighted the need for further NHI training and communication. Our findings suggest that such efforts should focus particularly on the practical mechanisms through which NHI will be implemented rather than only on its broad objectives.

Building on these descriptive findings, the repeated-measures ANOVA revealed an important distinction between support for the reform and confidence in its implementation. Participants were similarly positive about NHI and its potential effectiveness, but less confident about readiness/capacity and substantially more concerned about implementation risks. Workforce adequacy, governance, payment, supply-chain processes, and workload were prominent concerns. This finding is consistent with those of Mukwena et al. [10] and Mudzweda et al. [8], who reported constraints related to workforce, infrastructure, resource and preparedness for NHI rollout. Our study extends this evidence by showing that such concerns can coexist with relatively favourable attitudes towards NHI and confidence in its potential effectiveness. This distinction also supports Mofolo and Heunis [9] warning that systemic constraints may prevent policy ambition from becoming sustainable implementation.

The correlation analysis helps explain this apparent coexistence of support and concern. Greater knowledge was associated with more favourable attitudes and stronger perceptions of NHI effectiveness, but not with greater confidence in readiness; it was also associated with greater implementation concern. Greater knowledge therefore did not correspond to uniformly more positive perceptions. Rather, more knowledgeable professionals appeared to distinguish between the desirability of NHI and the system's capacity to deliver it. Compared with previous studies that largely examined NHI awareness, attitudes or readiness separately, a strength of the present study is its simultaneous examination of these dimensions and their interrelationships. Greater knowledge may therefore foster a more informed and differentiated view of NHI rather than simply greater optimism.

This differentiated picture was also reflected in the MANOVA findings. Professional differences were concentrated in knowledge and readiness rather than attitudes, while older participants tended to report more favourable attitudes and greater confidence in readiness. Prior exposure to NHI information was associated with greater knowledge, but not with attitudes or perceptions. Consistent with the training and communication needs identified by Mudzweda et al. [8], this suggests that exposure to information may improve knowledge without necessarily changing how healthcare professionals evaluate the reform or its implementation.

The regression analysis added a final layer to this pattern. Knowledge remained the strongest independent predictor of attitudes after profession, age, experience and prior exposure were considered. Interpreted alongside the correlation findings, however, this does not imply that greater knowledge produces uncritical support; instead, more knowledgeable professionals also demonstrated greater awareness of potential implementation risks. Age and experience remained relevant, suggesting that attitudes may reflect both factual understanding and accumulated professional experience.

Taken together, these findings suggest that implementation readiness should not be inferred from favourable attitudes towards NHI alone. Healthcare professionals may support the objectives of the reform while remaining unconvinced that the organisational conditions for successful implementation are in place. This has implications for both policy and practice. National health insurance communication should move beyond general awareness-raising towards clearer explanation of how the reform will operate in practice, while policymakers and healthcare organisations should concurrently address concerns about workforce capacity, governance, payment systems, supply chains and workload. Involving healthcare professionals directly in local readiness assessment and implementation planning may be particularly valuable. Their concerns, especially when expressed by more knowledgeable frontline professionals, should not automatically be interpreted as resistance but may provide useful evidence about weaknesses that need to be addressed before and during implementation. These findings also point to unanswered questions. Future research should examine whether the same relationships are found across other public healthcare facilities and professional groups. Multi-site and longitudinal studies could further explore whether changes in NHI knowledge and health-system readiness are accompanied by changes in healthcare professionals' attitudes as implementation progresses.

Limitations: several limitations should be considered when interpreting these findings. The cross-sectional design does not permit causal conclusions, although establishing causality was not the purpose of this exploratory study, but rather to provide insight into the knowledge, attitudes and perceived readiness of healthcare professionals within the participating public hospital. This, in turn, could potentially inform a larger programme of research. The single-hospital setting and predominantly female nursing sample limit the extent to which the findings can be generalised; the specific readiness concerns identified should therefore be applied cautiously to other hospitals or districts. Compared with studies conducted across multiple facilities or districts, the present study consequently provides a narrower organisational perspective. Attitudes and perceptions were also self-reported and may have been influenced by social desirability bias. However, the substantial implementation concerns expressed indicate that respondents reported a range of critical views regarding NHI, although the possibility of social desirability bias cannot be entirely excluded. Concerning recall bias, responses may also have been influenced by participants' ability to accurately recall previous exposure to NHI-related information. These limitations are balanced by a multidisciplinary sample, the integrated examination of several NHI-related constructs, and study scales with acceptable-to-excellent internal consistency.

Conclusion     

This study shows that support for NHI and concern about its implementation can exist side by side. Greater knowledge was associated with more favourable attitudes towards NHI, but also with greater awareness of implementation risks. Successful implementation will, therefore, require more than increasing awareness: it will require targeted communication, informed engagement with healthcare professionals and effective measures to respond to practical concerns raised about readiness. As NHI implementation progresses, strengthening operational knowledge while addressing workforce, governance and organisational constraints will be important for translating support for the reform into confidence in its implementation.

Competing interests     

The authors declare no competing interests.

Authors' contributions     

Virginia Apandju Osako Ngende conceptualized and developed the first draft of the manuscript. Subsequently, Burt Davis intellectually revised, guided and contributed to the drafted manuscript. Virginia Apandju Osako Ngende conducted the data collection process. Both authors were involved in the analysis process. Both authors have read and approved the final version of this manuscript.

Acknowledgments     

The authors would like to thank Stellenbosch University Ethics Committee for approving the study and Ekurhuleni Health District Research Committee for granting permission to conduct the study within the district. The authors would like to extend their gratitude to the executive management team for authorizing the study to be conducted in the hospital. In addition, thank you to all the hardworking and dedicated healthcare professionals who volunteer to participate in the study. Moreover, special thanks to Prof Marie-Claire Omanyondo for her intellectual support and contribution. The content of this paper is, however, the responsibility of the authors.

Tables     

Table 1: demographic and professional characteristics of the study participants (N = 203)

Table 2: descriptive statistics for national health insurance knowledge, attitude, and perception

Table 3: Pearson correlations among national health insurance knowledge, attitude, and perception domains

Table 4: hierarchical multiple regression predicting attitude towards national health insurance

References