Task shifting can be a potential solution to mitigate the challenges of brain drain and improve the quality of service delivery in healthcare institutions.¹⁹ Therefore, the aim of this study was to assess the efficacy of task-shifting interventions for mental health in Zimbabwe. The findings could assist the development and implementation of support programmes to help healthcare practitioners use task-shifting and task-sharing as a strategy implemented by psychiatric hospitals.

This study utilised a descriptive cross-sectional survey design to assess perceptions of the effectiveness, implementation and implications of task shifting in a psychiatric hospital in Zimbabwe. The cross-sectional design enabled comparison of multiple variables at a single point in time and was appropriate for evaluating associations without inferring causality.

The study was conducted at a psychiatric hospital in Bulawayo, Zimbabwe. The institution provides a wide range of mental health services, including acute, subacute, forensic, rehabilitation and psychogeriatric care. The facility has a bed capacity of 708, with an average daily inpatient population exceeding 600 and an outpatient department catering to an average of 2400 patients per month.

The study population consisted of HCWs at a psychiatric hospital. The inclusion criteria were HCWs whose roles involved being recipients of shifted task delegation, including nurses, doctors and nurse aides. Exclusion criteria included HCWs who were not delegated shift tasks, temporary HCWs on locum contracts and individuals on short-term academic attachments at the hospital.

The estimated sample size was 52, calculated using the Centers for Disease Control and Prevention’s Epi Info statistical software, as shown in Figure 1.

Figure 1 illustrates how the sample size varies with the confidence level when conducting a population survey using simple random sampling. It uses a standard formula for estimating sample size for proportions, based on expected frequency, margin of error and population size. The calculator estimates the number of individuals the researchers need to survey from a population of 502 HCWs, assuming an expected frequency (prevalence) of 8%, which is the proportion of the population expected to exhibit the trait or condition. In Figure 1, the acceptable margin of error is ± 7%, the design effect is 1.0, and it assumes simple random sampling, with clusters of 1, indicating no clustering. To manually sample, the researchers employed a systematic sampling technique, where they selected the HCW at every tenth name on the randomly ordered population list. A total of 53 participants completed the questionnaire, providing a 100% usable response rate.

Data were collected using a structured online questionnaire designed and distributed via Google Forms. It consisted of multiple-choice questions, Likert-scale items, checkbox items and short responses. Content validity was ensured through expert review and pilot testing, achieving a content validity index (CVI) of 0.75. Test–retest reliability was assessed, yielding a high internal consistency with a correlation coefficient of 0.978.

The Google Forms research questionnaire tool supported automatic data collection and real-time response monitoring, allowing the researcher to track participation and analyse data promptly. Furthermore, the ability to distribute the questionnaire via email, WhatsApp Messenger or embedded links broadened the reach, ensuring a diverse and representative sample.

The questionnaire was pre-tested through a pilot study to ensure its validity and reliability. The pilot questionnaire was sent to HCWs in a setting similar to the anticipated study environment. Participants provided feedback on the content and layout of the questionnaire.

Approximately 67% of the participants expressed concerns about the length and complexity of some questions, suggesting that they could be summarised and worded better. The layout was generally well-received, but participants recommended more information on the purpose and scope of the questionnaire in the introduction. Responses were systematically organised in a Google Sheet, facilitating straightforward data manipulation and analysis.

Multiple strategies were employed to ensure the validity and reliability of this task-shifting study. Validity was addressed using a well-structured and pre-tested questionnaire aligned with the research objectives and theoretical frameworks established in previous studies. The questionnaire underwent a pilot test to identify and rectify any ambiguities, ensuring that questions were clear and relevant to the study’s aims. External validity was enhanced by employing a representative sample of HCWs in a public healthcare setting similar to the research setting, enabling the generalisation of the findings to the broader population. Reliability was ensured through standardised procedures for data collection, including training data collectors to ensure uniformity in administering the questionnaire and recording responses.

Test-retest reliability was assessed by administering the same questionnaire to a sample subset at two different points in time and calculating the correlation between the two sets of responses.

Data collection was conducted electronically. The questionnaire link was distributed to participants via institutional email and WhatsApp. Google Forms supported real-time monitoring, automatic collation and secure storage of responses. Participants completed the survey at their convenience, reducing response burden.

Data were exported to Microsoft Excel for initial cleaning, coding and verification. This process involved checking for the accuracy, completeness and consistency of the responses. Statistical analyses were performed using SPSS and Epi Info.

Descriptive statistics were used to summarise demographic characteristics and task-shifting variables in order to illustrate the findings. Inferential statistics and their claims, complement the study, as they attempt to generalise beyond the sample.²⁰ In our study, the effect sizes indicate the magnitude of task shifting, while the confidence intervals indicate the level of certainty. Together, they provide a richer interpretation than p-values alone, ensuring that both descriptive and inferential claims are properly contextualised and understood. Inferential statistics included Chi-Square tests to assess associations, correlation analysis to evaluate relationships between scope clarity and task ambiguity and logistic regression to identify predictors of perceived quality of care. No causality was inferred because of the cross-sectional design.

The researchers adhered to all ethical principles regarding human subjects, and ethical approval was obtained from the Research and Ethics Committee of the University of Johannesburg, South Africa, prior to data collection. The ethical clearance number is REC-2247-2023. The study was conducted in accordance with ethical guidelines, ensuring that informed consent was obtained from all participants. This was achieved by including an informed consent statement on the first page of the form, which required participants to indicate their consent before accessing the questionnaire. Participation was voluntary, and informed consent was obtained from all participants. Confidentiality was maintained by anonymising all data, and responses were stored securely to prevent unauthorised access.

Fifty-three HCWs participated in the study. Most participants were aged 25–44 years, and the majority were female. Professional diplomas were the most common qualification. The age distribution differed significantly from uniformity (p < 0.001), indicating a predominantly mid-career workforce. Educational level was significantly associated with task delegation patterns (p < 0.01). Gender and profession were not significantly associated with delegation patterns (see Table 1 for demographics and educational qualifications).

As visualised in Table 1, the age distribution was not uniform (χ² = 212, p < 0.001), with the largest proportions of participants in the age groups 35–44 years (35.85%) and 25–34 years (28.30%), suggesting a workforce skewed towards mid-career professionals.

Gender distribution showed a non-significant deviation from uniformity (χ² = 2.15, p = 0.709), with female participants representing 61.54% of the sample. Though numerically more prevalent, the difference was not statistically significant.

Educational attainment was significantly concentrated among holders of a professional diploma (77.36%; χ² = 37.65, p = 0.0098), suggesting a dominant educational background in diploma-level training.

Regarding profession, the most frequently reported role was psychiatric nurse (35.85%). However, the chi-square test indicated no significant deviation from uniform distribution (χ² = 22.26, p = 0.564), reflecting a relatively balanced professional representation across roles.

In terms of professional qualification, the data showed an over-representation of participants holding a diploma (83.02%), though the result did not reach statistical significance at the conventional 0.05 level (χ² = 29.47, p = 0.079). This finding nonetheless indicates a homogeneity in formal qualification levels within the sample.

Older HCWs were significantly more likely to perform untrained tasks (p = 0.0103). No significant associations were observed for gender, education or years of experience (see Table 2).

Additionally, while HCWs felt capable and supported in their task-shifting roles, there were significant concerns regarding mental health and the adequacy of training. Age was a significant factor in the likelihood of performing untrained tasks, with older HCWs more likely to engage in such activities. However, no significant associations were found between gender, years of experience, and the performance of untrained tasks (Table 2).

Task shifting was a significant predictor of perceived care quality (β = 4.48, p = 0.002). Participants with more than 20 years of experience reported lower perceived care quality (β = −2.42, p = 0.029), suggesting potential adaptation challenges or burnout. None of the other covariates, including professional qualification (diploma, degree, master’s, etc.) and years of experience, showed statistically significant associations (p > 0.05 for all), with the exception of those with more than 20 years of experience, who reported lower perceived quality (β = −2.42, p = 0.029) (see Table 3).

Thirty per cent of participants reported dissatisfaction with their jobs, 50% with their working conditions and 40% with task-shifting support structures. Clarity of scope of practice showed a strong positive correlation with perceived confidence (r = 1.0) (see Table 4).

Figure 2 shows participants with more than 20 years of experience having a lower median quality of care rating. This indicates that, contrary to what might be expected, extensive experience does not necessarily correlate with higher perceived quality of care in this sample. The study’s findings suggest that task shifting can have both positive and negative impacts on healthcare delivery. While it enhances workforce efficiency and service coverage, it also raises concerns about job satisfaction and working conditions.

Figure 3, on the other hand, illustrates the distribution of knowledge among the staff.

Knowledge of scope is generally high, although it is inconsistently reported. Thus, the availability of scope information showed a strong positive correlation with knowledge (r = 1.0) but a negative correlation (r = –0.25) with the perception that unavailability leads to undefined tasks. The perceived unavailability of scope information was moderately associated with dissatisfaction and concerns about task ambiguity, especially in contexts where formal guidelines are lacking.

A detailed illustration of the distribution of untrained task performance is shown in Figure 4.

The findings underscore the urgent need for policies that formalise task-shifting frameworks through standardised training curricula, clear supervision models and well-defined scopes of practice. Such formalisation would safeguard the quality of care and ensure that delegated responsibilities are performed safely and effectively.²³ Strengthening supervision structures is particularly critical, as evidence shows that consistent oversight enhances accountability and supports safe delegation practices.²⁷

Equally important are investments in continuing professional development, which can bolster HCWs’ confidence, improve performance and sustain motivation in resource-constrained settings.²⁵ Improving working conditions and embedding mental health support systems for HCWs should also be prioritised, given the documented risks of burnout, emotional strain and demoralisation among frontline staff engaged in task-shifting roles.²⁴

Taken together, these implications highlight that task shifting is not merely a stopgap measure but a strategic health systems intervention. When supported by robust training, supervision and workforce wellbeing initiatives, task shifting can enhance service delivery, optimise specialist expertise and strengthen the resilience of health systems facing chronic human resource shortages.²²

A key strength of this study lies in the use of systematic sampling, which enhanced representativeness and reduced selection bias, thereby strengthening the validity of the findings.²⁸ The application of validated data collection tools with demonstrated reliability further ensured consistency and accuracy in measurement.²⁹ In addition, the study employed appropriate statistical methods for a cross-sectional design, allowing for robust analysis of associations within the dataset.³⁰

However, several limitations must be acknowledged. Firstly, the reliance on self-reported data introduces the possibility of recall bias and social desirability bias, which may affect the accuracy of responses.³¹ Secondly, the single-site setting limits external validity and may reduce the generalisability of findings to other contexts, particularly across diverse health systems in Africa. Thirdly, the cross-sectional natureof the study precludes causal inference, restricting conclusions to associations rather than directional effects.³²

Future research should therefore adopt multi-site designs to capture broader contextual variations and enhance generalisability. Moreover, mixed-methods approaches could provide deeper insights into the implementation challenges and workforce perceptions surrounding task shifting, combining quantitative rigour with qualitative exploration of lived experiences.³³ Such designs would strengthen the evidence base and inform policy frameworks for scaling task-shifting strategies in mental healthcare.