Skip to main content

Measuring productivity and its relationship to community health worker performance in Uganda: a cross-sectional study

Abstract

Background

To explore the nature of the relationship between and factors associated with productivity and performance among the community health volunteer (CHV) cadre (Village Health Teams, VHT) in Busia District, Eastern Uganda. The study was carried out to contribute to the global evidence on strategies to improve CHV productivity and performance.

Methods

This cross-sectional study was conducted with 140 VHT members as subjects and respondents. Data were collected between March and May 2013 on the performance and productivity of VHT members related to village visits and activities for saving maternal and child lives, as well as on independent factors that may be associated with these measures. Data were collected through direct observation of VHT activities, structured interviews with VHTs, and review of available records. The correlation between performance and productivity scores was estimated, and LASSO regression analyses were conducted to identify factors associated with these two scores independently.

Results

VHTs demonstrated wide variation in productivity measures, conducting a median of 13.2 service units in a three-month span (range: 2.0-114.9). Performance of the studied VHTs was generally high, with a median performance score (out of 100) of 96.4 (range: 50.9-100.0). We observed a weak correlation coefficient of 0.05 (p = 0.57) between productivity and performance scores. Older VHT age (≥50 years old, reference: <50 years old) (11.14, 95% CI: 3.26-19.01) and knowledge of danger signs (in units of ten-percentage points, 1.92, 95% CI: 0.01-3.83) were positively associated with productivity scores. Job satisfaction (1.46, 95% CI: 0.13-2.80) and knowledge of danger signs (in units of ten-percentage points, 1.02, 95% CI: 0.05-1.98) were positively associated with performance scores.

Conclusions

Older VHT age and knowledge of danger signs were positively associated with productivity, and job satisfaction and knowledge of danger signs were positively associated with performance. No correlation was observed between productivity and performance scores. This lack of correlation suggests that interventions to improve CHV effectiveness may affect the two dimensions of effectiveness differently. We recommend that productivity and performance both be monitored to evaluate the overall impact of interventions to increase CHV effectiveness.

Peer Review reports

Background

Inspired by the Alma Ata health care for all movement, community health worker (CHW) / community health volunteer (CHV) programs have been established in many countries to empower communities to reduce preventable mortality and morbidity. A number of studies and high-level forums have recognized the essential contributions that CHVs and CHWs make and their impact on the health of the community [1,2,3,4]. CHW and CHV effectiveness is affected by how well they do their work against expected standards (performance), how much they can feasibly do (productivity), and what factors affect their work [5]. Performance is defined in the literature in a multitude of ways, and some definitions incorporate the element of productivity. For our purposes, performance is defined as the extent to which tasks or services are completed according to standards, as a measure of how well tasks are done; whilst productivity is defined as the volume of services, or how much is done. A better understanding of these factors will allow the greater health community to deploy and support them to maximize their effectiveness.

The expectations of CHWs and CHVs have increased over time, with many countries expanding their scope of practice and responsibilities, potentially overburdening them [6]. CHW and CHV performance has been studied extensively [5, 7, 8], but fewer studies have examined their productivity [5, 9]. To our knowledge, the interplay of factors that can affect both productivity and performance has not been examined in detail.

In the year 2000, the Ugandan Ministry of Health (MoH) established Village Health Teams (VHT) to empower communities to take part in the decisions that affect their health, to mobilize communities for health programs, and to strengthen the delivery of health services at the household level. VHTs are unsalaried CHVs who are linked to the lowest level of the health care system, or Health Center I. The MoH reports the following guidelines for VHT selection: must be 18 years of age or older, reside in the village, literate in the local language, capable of community mobilization and communication, dependable, trustworthy, interested in health and development, and willing to work for the community [10]. The MoH recommendation is that one male and one female community member be selected through popular vote of stakeholders and households in the village to constitute the VHT, though in larger communities, more than two VHT members may be chosen.

A VHT is responsible for approximately 30 households [11]. VHT responsibilities are delineated into seven categories: record (collecting data on catchment population), model (practicing model health behaviour), save lives (providing health services and referrals), link (improving linkages between formal health care and their catchment areas), mobilize (mobilizing village members and health workers for health activities), visit (visiting households and communities), and support (assisting in care of those living with illness / HIV) [12]. More details of their responsibilities are available in Additional file 1. In recent years, VHTs have been increasingly tasked with strengthening community support for people living with HIV, including tasks such as mapping HIV patients in the community; linking patients to VHTs and other resources; engaging community groups to support HIV patients to address their needs such as nutrition, psychosocial support, savings, and loan; and supporting patient self-care through health goal setting to improve health status [13]. Several studies have demonstrated the importance of VHTs in Uganda and their potential to further improve health services and outcomes [14, 15].

This study seeks to describe VHT productivity and performance in Busia District, Uganda, explore the relationship between VHT productivity and performance, and investigate independent factors that influence VHT productivity and performance. The study was conceived to contribute to the global evidence on strategies to improve CHW/CHV productivity and performance with an eye toward better understanding factors that could increase their effectiveness.

Methods

Study design

We conducted a cross-sectional study to assess levels of VHT productivity and performance in Busia District, Eastern Uganda between March and May 2013. Training of the VHTs was conducted by the MoH in collaboration with World Vision International using the standard training guide for VHTs developed by the MoH. Training length varied though the same training package and facilitators guide was used. Specific details of training undertaken by each VHT were not available. Supervisory systems varied across districts. In the district under study, the VHTs were supervised by World Vision International through quarterly review meetings that brought all the VHTs together.

To quantify productivity and performance, our study focused only on the “save lives” and “visit” tasks of the basic package of services as defined by the Ugandan MoH (full description of the six components of the basic package of services can be found in Additional file 1). The study population included all VHTs who provided the basic package of services in two randomly selected sub-counties out of a total of ten sub-counties in Busia District. Data were collected through multiple sources. In one, the data collector observed a VHT assess a sick child (2-59 months) and a pregnant mother for danger signs during a home visit. The VHT recorded his/her findings, illness classification, and management on an observation checklist. Immediately after the consultation, the data collector conducted a comparative, “gold standard” clinical evaluation of the child to determine 1) if the VHT had missed any danger signs that required immediate attention or referral and 2) the extent to which the VHT’s evaluation was correct against the standard. During the same visit, the quality of interaction with the household and the length of time the home visit took, including travel time, were recorded as well. For quality of the interaction, a checklist was used by the data collectors that included points such as whether the VHT addressed the family politely and respectfully, whether relevant advice was given, and whether the VHT facilitated agreement with the family on what actions should be taken by whom and when. Data collectors were instructed to build rapport with the observed VHTs to minimize any issues. Due to resource limitations and logistical constraints, only one direct observation was conducted per VHT.

The second data source was routine data already being collected by VHTs, such as village maps, village registers, and monthly summary forms. Monthly data were taken from August to October 2013. These were reviewed for completeness and to quantify the frequency of activities conducted by the VHTs. The frequency data were taken from VHT registers and monthly summary forms. Finally, surveys were administered to VHTs and supervising health facility staff to collect data on individual characteristics of the VHTs. The variables chosen for collection were identified from the literature and also suggested by stakeholders in a November 2012 consultation held in Kampala, Uganda, and grouped into categories (Additional file 1). Data were triangulated where possible to address discrepancies, and some data were excluded because they formed patterns of reporting that were not feasible. Occasionally, frequency of a particular intervention was not documented, thus self-report was used. This was then verified with those interviewed in the village to ensure that these were as accurate as possible.

Data collection tools were translated from English into the local language in Busia, independently reviewed and revised, and back-translated into English for comparison. They are available on Additional file 2. Data collection tools were piloted in a sub-county not selected for the study. Data were entered into a proforma template in EpiInfo. Data regarding reported frequency of activities were cross-checked with standard documentation such as monthly reports and VHT registers, where available. Data entry quality checks were completed for 10% of the data.

Sample size

The following equation was used for sample size calculation: n = [DEFF*Np(1-p)]/ [(d2/Z21-α/2*(N-1) + p*(1-p)] [14]. The sample size formula is standard for prevalence-based sampling. We used the VHT’s ability to correctly elicit and respond to danger signs as basis for the sample size calculation, using 67% (p) as reported in a 2011 study in the Eastern Districts of Uganda [14]. We assumed this percentage applied to Busia District. Note that there was no other comparable basis upon which to calculate the sample size. With a precision of 5% (d) with a 95% confidence interval (α of 0.05), and a design effect (DEFF) of one, we estimated that each sub-county needs a sample of 65 VHTs to detect the 67% prevalence.

Ethics, consent, and permission

Ethics approval was received from the Institutional Review Boards of University Research Co., LLC, Makerere College of Health Sciences, and the Uganda National Council for Science and Technology. Verbal informed consent was obtained from all participants.

Productivity and performance measures

Productivity

Productivity was measured as the total volume of output of a service provided by one VHT during the three months prior to the date of data collection. The study collected data on the frequency of completion of the activities/tasks related to visiting villages and assessing danger signs of children and pregnant mothers, and then aggregated them into a composite unit that represented a single volume of output. The middle column of Table 1 describes the variables used to derive the composite productivity score.

Table 1 Description of VHT productivity and performance measures

We adapted a method used by the World Bank to create a Composite Services Index (CSI) [16], which is an aggregate measure of the quantity of health services delivered by health workers. CSI is expressed as service equivalents. For our purposes, we used the unit of “home visit” as the service to make the other services (evaluation of children and pregnant mothers) comparable to. The details on the derivation of CSI are available in Additional file 1. The CSI served as the productivity score.

Performance

Performance was calculated as a standardized score, representing the extent to which tasks were completed to expected standards. We collected data on the quality of interactions between VHT and families during home visits (task two of basic package of services) and on how accurately VHTs identified danger signs in pregnant women and children (task three of basic package of services). The right column of Table 1 describes the variables used to derive a performance score. Each cell of performance measures listed in the right column of Table 1 was standardized into a scale of 0-100%, averaged for each task group first, then averaged again to derive the performance score.

Data analysis

We calculated and summarized the productivity and performance scores and the independent variables (listed in Table 2) across VHTs. The correlation coefficient was calculated between the productivity and performance scores to describe the relationship between the two variables. We then ran regression analyses to estimate the associations between the independent variables and productivity and performance, respectively. Independent variables that we were able to collect complete data on and had enough variability are listed in Table 2 and were included as covariates in the regression models. For pre-defined groupings of independent variables, non-linear Principal Component Analysis (PCA) was conducted to reduce the number of variables. The first component, which explains the largest amount of variation among the selected variables, was retained as the summary measure for that set of variables [17, 18]. The categories that summary measures were created for include VHT’s job satisfaction, VHT’s belief that he/she is a change agent, VHT’s sense of accountability, and how VHTs perceived the quality of the supervision they received. Equipment availability was tabulated, out of twelve total pieces of equipment identified as essential or necessary for VHT activity (mid-upper arm circumference tape, VHT handbook, job aids, referral forms, summary forms, VHT register, VHT maps, respiratory timers, soap, clean water, thermometer, and information, education, and communication materials).

Table 2 Description of factors that may influence productivity and performance

Multi-level multiple linear regression, accounting for clustering at the sub-county level, was conducted to identify variables that are statistically significantly associated with productivity and performance independently. Separately, LASSO regression [19] was also conducted; this statistical method takes into account the collinearity between variables and creates a more parsimonious model through variable selection. The selected variables were refitted in a separate multi-level multiple linear regression model to account for regression coefficients that are biased toward zero in the LASSO regression [20, 21]. Stata version 13.0 (StataCorp, College Station, TX) was used for the analysis. The data file is available upon request.

Results

Data were collected for 147 VHTs available in these two districts, 79 from Sub-county 1 and 68 from Sub-county 2. Five VHTs had substantial missing data, and two VHTs had outlier quantities in key variables comprising the productivity variable. A final total of 140 (95.2% of total surveyed) VHTs were thus included in the analysis. Their background characteristics are described in Table 3. Of the 140 VHTs, 52% were female and 51% were between the ages of 35- < 50. The VHTs had an average of 9.3 years of education and an average age of 40.7. The loadings for the first component from the PCA-derived variables are available in Additional file 1.

Table 3 Characteristics of VHTs included in the study, % or mean (SD)

The productivity and performance scores are summarized in Table 4. CSI, with each unit an equivalent of a home visit, ranged widely. VHTs conducted a median of 13.2 and mean of 20.8 service units in a three-month span, with a range of 2.0-114.9. The performance score was generally high, with a mean of 91.0 and median of 96.4, with a range of 50.9-100.0. There was no evidence of either positive or negative correlation between productivity and performance, with a correlation coefficient of 0.05 (p = 0.57).

Table 4 Summary of productivity and performance measure

The results from the full multiple linear regression models are available in Additional file 1. We present the results for the regression models containing the retained variables from the LASSO regression in the text and in Table 5. For productivity, older VHT age and VHT’s knowledge of danger signs were significantly associated with greater productivity at the p < 0.05 level; other VHT characteristics, such as level of engagement and competency, were not found to be significantly associated. VHTs who were 50 years or older had greater productivity than their counterparts younger than 50 years old (11.82, 95% CI: 4.03, 19.60), and for every ten-percentage point increase in knowledge of danger signs, the productivity score increased by 1.92 (95% CI: 0.01, 3.83). Years serving as a VHT had a marginally significant positive association with productivity (0.81, 95% CI: -0.05, 1.67) (Table 5). For performance, job satisfaction and knowledge of danger signs were significantly associated at the p < 0.05 level. For every unit increase in job satisfaction, the performance score increased by 1.46 (95% CI: 0.13, 2.80) and for every ten-percentage point increase in knowledge of danger signs, the performance score increased by 1.02 (95% CI: 0.05, 1.98). At the p < 0.10 level, mothers who were < 18 years old had VHTs whose performance was 5.62 units lower than mothers aged ≥18 (95% CI: -12.14, 0.90) and for every additional year of maternal education, the VHTs performed worse by 0.53 percentage points (95% CI: -1.08, 0.02) (Table 5).

Table 5 Covariates associated with productivity and performance, selected through LASSO regression

Discussion

CHVs serve a critical role in providing life-saving health interventions in low-resource settings, particularly among the hardest-to-reach populations. Productivity and performance are both key measures in assessing how effectively CHVs are serving their respective communities. We observed in our study that there is minimal correlation between measures of VHT productivity and performance in two sub-counties of Busia District, Uganda, and we only observed that knowledge of danger signs were significantly associated with both measures. This implies that while there was no evidence that performance was being sacrificed for increased productivity or vice versa, there was also no evidence to suggest that improvements in one would contribute to improvements in the other, or that there are specific characteristics of VHTs or their environments that could be targeted to improve both. This highlights the need for conscious policy planning that recognizes that these two components of VHT effectiveness may need potentially integrated, but nonetheless independent, interventions. Also, VHTs require ongoing practical training and supportive supervision to develop and maintain competencies in identifying danger signs in pregnant mothers and children. We believe our findings are generalizable to the greater VHT system in Uganda and to similar CHV models.

Factors affecting productivity and performance

A desk review on health worker productivity highlighted knowledge and skills, motivation, and the work environment (workload, supportive supervision, supplies and equipment, respect) as determinants of productivity [5]. We were not able to find a statistical relationship between productivity and those characteristics that we were able to capture in our study. This may be attributable to several factors. One, contextual differences may affect how influential each factor is on productivity. Dieleman et al. reported that health resource management interventions on CHW performance have shown different effects in different contexts [22], and the same issue likely applies to productivity. We may not have successfully measured or captured the factors that influence productivity specifically among the CHVs in our context. Two, certain factors may be modifying the relationship of other factors to productivity. Due to our small sample size, we were unable to examine interactions between various independent factors. In the existing literature, a study comparing interventions that focused on CHW training noted that the selection process of the CHWs (i.e., by the beneficiary community themselves vs. by the political establishment) and the targeted need of the community (unmet vs. met need) were contextual factors that modified the intervention outcome [7]. Similar analyses on effect modification should be conducted in the future with productivity as an outcome, as much of the existing literature focuses on performance.

VHT characteristic of being older than 50 years of age (ref: 35- < 50 years of age) and knowledge of danger signs were statistically significantly associated with productivity. Independent from age, number of years of serving as a VHT was marginally significant. Engagement, incentives, recognition, availability of equipment, and supervision were not found to be associated with productivity in this study. While we were unable to identify existing studies examining the relationship between CHV age and productivity, other studies have shown mixed findings regarding health worker’s age and performance [23]. A study in India reported that absenteeism was highest among women aged 25-40 due to pregnancies and other maternal health issues [24]. Another study in Kenya, examining age groups ranging from a “below 30” category to an “above 60” category, saw highest performance among women aged 40-50 years [25]. The authors attributed this to social stability within that age group. It is also important to note that experience may not consistently correlate with improved performance or productivity; a study examining the performance of health workers on neonatal resuscitation in Malawi saw that health workers with longer years of experience showed the poorest levels of performance. That study hypothesized that newer standards of care may need reinforcement among individuals who have more experience [26]. The fact that older mothers may prompt better quality performance is consistent with other literature [27]. The additional years of maternal education was associated with poorer CHV performance, however the effect size is negligible and it is unclear what the operating mechanism would be.

The magnitude of effect sizes was quite small for performance, thus whilst some indicators were statistically significantly associated, they did not appear to have a meaningful bearing on performance. Job satisfaction was a factor significantly associated with performance, however not with productivity; previous studies have described the relationship with performance as well. Several studies from different contexts have highlighted low salary and lack of professional development opportunities as key components influencing job satisfaction [28,29,30]. The salary issue is notable, as financial incentives have been reported in certain situations to have a negative impact on motivation [22]. This distinction between salary and financial incentive may be reflective of CHWs valuing respect and formalized support from the government and the health system. In a separate study on VHT motivations, VHTs expressed a desire to receive a stipend that is approximately 75% of the minimum public sector wage and for the VHT role to become something that resembles a full-time job [31]. A third of the VHTs indicated that their work negatively affected household income. Many of them expressed willingness to take on more work than the current workload, but it is unclear if and when the cost-benefit ratio will shift to deter volunteering their time, as the government utilizes VHTs for additional tasks. Using several different methodologies to calculate opportunity cost of CHV time used toward their work, the authors of the study recommended a monthly stipend of 5-10 USD for CHVs to cover basic expenditures and receive a commensurable reward. The lowest salary level within the Ugandan public health sector was 54 USD per month at the time of the study [31].

The Ugandan MoH is moving toward a salaried CHW program. A national assessment of 2610 VHTs in Uganda conducted in 2015 found that across the country, forms of monetary and non-monetary motivation for VHTs have been irregular, varying across implementing partners and districts, undermining coordination and retention of VHTs. While the government’s VHT Operational Guidelines stipulate the various forms of motivation for VHTs, it does not explicitly state how these incentives should be equitably distributed and who should provide them. Qualitative data from VHT interviews overwhelmingly supported regular financial form of motivation. Such regular financial incentives were advocated for in the Ministry’s 2015 Community Health Extension Workers Strategy which recommends replacing the VHT program with full-time salaried community health extension workers who would be government employees [32]. The true impact of financial compensation on job performance, retention, and sustainability has not been properly examined in many contexts; Uganda’s transition may provide a unique opportunity to assess how compensation affects CHV or CHW programs.

One of the strengths of our study is that we were able to conduct direct observations to assess key productivity and performance measures. Many of the prior published studies of CHV performance have been done in health facilities or in situations that do not represent the context within which they provide their services. The observations allowed for the validation of danger sign assessment and for measurement of the length of a home visit, permitting quantification of productivity. There has been documented overestimation of performance during observations, or the “Hawthorne effect”, compared to situations where CHVs or CHWs were unaware that they were being observed. However, without direct observation, performance on danger sign assessment could not be compared and interviews would be subject to biased responses.

Limitations

While our study focused on the performance and productivity of CHVs, guaranteeing high performance and high productivity alone does not necessarily lead to improved health outcomes. We only report the CHW-level side of health care access and do not address the end-user in our study, and as such were unable to determine the relationship between CHV productivity or performance and health outcomes, as that went beyond the scope of the study. Socioeconomic and cultural barriers to referral completion have been discussed extensively [33,34,35], and high quality of care at referral facilities is not guaranteed. For severe cases in which referral completion is vital to saving a life, the factors beyond the first point-of-contact with the CHV will need to be addressed before CHV programs can maximize their utility. There appeared to be variability in productivity levels, but we did not have sufficient information to understand the variation.

The cross-sectional study design does not allow us to determine directionality of the associations we observed here. For instance, we observed a negative association between supervisory visits and performance. This may be due to worse-performing VHTs being targeted for additional supervisory visits. Also, we were unable to collect complete data on the full set of tasks in the VHT basic package of services, thus we could not capture productivity and performance scores that reflect the entirety of a VHT’s responsibilities. This limitation may have led to an under-estimation of VHTs’ productivity. Finally, with our small sample size, the limited variability in the dependent and/or the independent variables may have prevented us from capturing some of the associations.

Conclusion

We observed no correlation between productivity and performance among VHTs in Eastern Uganda. We also did not identify variables that were associated with both measures. We found that older VHT age and knowledge of danger signs were positively associated with productivity and that job satisfaction and knowledge of danger signs were positively associated with performance. The lack of correlation between productivity and performance in this study suggests that interventions to improve CHV effectiveness may affect each of these two dimensions of effectiveness differently. While our study did not suggest that productivity gains must necessarily be sacrificed for performance gains, CHV improvement interventions, particularly for counselling and support tasks associated with community-level services, may not affect both. CHW interventions include improved clarity in roles, competency development, and supportive supervision and feedback [8, 36]. We recommend that both dimensions of CHV effectiveness be monitored to evaluate the overall program impact of CHV improvement interventions. Consideration of the independent effects of improvement interventions on productivity and performance is likely to be relevant even as the Government of Uganda considers moving away from a CHV to a paid CHW model.

Abbreviations

CHV:

Community health volunteer

CHW:

Community health worker

CI:

Confidence interval

CSI:

Composite Services Index

DEFF:

Design effect

HIV:

Human immunodeficiency virus

PCA:

Principal Component Analysis

USD:

United States dollar

VHT:

Village Health Team

References

  1. Christopher JB, Le May A, Lewin S, Ross DA. Thirty years after Alma-Ata: a systematic review of the impact of community health workers delivering curative interventions against malaria, pneumonia and diarrhoea on child mortality and morbidity in sub-Saharan Africa. Hum Resour Health. 2011;9:27.

    Article  PubMed  PubMed Central  Google Scholar 

  2. Haines A, Sanders D, Lehmann U, Rowe AK, Lawn JE, Jan S, Walker DG, Bhutta Z. Achieving child survival goals: potential contribution of community health workers. Lancet. 2007;369(9579):2121–31.

    Article  PubMed  Google Scholar 

  3. Mwai GW, Mburu G, Torpey K, Frost P, Ford N, Seeley J. Role and outcomes of community health workers in HIV care in sub-Saharan Africa: a systematic review. J Int AIDS Soc. 2013;16:18586.

    Article  PubMed  PubMed Central  Google Scholar 

  4. Campbell J, Admasu K, Soucat A, Tlou S. Maximizing the impact of community-based practitioners in the quest for universal health coverage. Bull World Health Organ. 2015;93(9):590–590A.

    Article  PubMed  PubMed Central  Google Scholar 

  5. Jaskiewicz W, Tulenko K. Increasing community health worker productivity and effectiveness: a review of the influence of the work environment. Hum Resour Health. 2012;10:38.

    Article  PubMed  PubMed Central  Google Scholar 

  6. Perry HB, Zulliger R, Rogers MM. Community health workers in low-, middle-, and high-income countries: an overview of their history, recent evolution, and current effectiveness. Annu Rev Public Health. 2014;35:399–421.

    Article  PubMed  Google Scholar 

  7. Kane SS, Gerretsen B, Scherpbier R, Dal Poz M, Dieleman M. A realist synthesis of randomised control trials involving use of community health workers for delivering child health interventions in low and middle income countries. BMC Health Serv Res. 2010;10:286.

    Article  PubMed  PubMed Central  Google Scholar 

  8. Naimoli JF, Frymus DE, Wuliji T, Franco LM, Newsome MH. A community health worker "logic model": towards a theory of enhanced performance in low- and middle-income countries. Hum Resour Health. 2014;12:56.

    Article  PubMed  PubMed Central  Google Scholar 

  9. Sebastian MS, Lemma H. Efficiency of the health extension programme in Tigray, Ethiopia: a data envelopment analysis. BMC Int Health Hum Rights. 2010;10:16.

    Article  PubMed  PubMed Central  Google Scholar 

  10. Turinawe EB, Rwemisisi JT, Musinguzi LK, de Groot M, Muhangi D, de Vries DH, Mafigiri DK, Pool R. Selection and performance of village health teams (VHTs) in Uganda: lessons from the natural helper model of health promotion. Hum Resour Health. 2015;13:73.

    Article  PubMed  PubMed Central  Google Scholar 

  11. Geoffrey B, Lorna M, Clare K. Village health team functionality in Uganda: implications for community system effectiveness. Sci J Pub Health. 2016;4(2):117–26.

    Article  Google Scholar 

  12. Village Health Team: A handbook to improve health in communities. http://www.healthynewbornnetwork.org/hnn-content/uploads/VHT-BOOK.pdf.

  13. Namwabira M, Nabisere N, Kisamba H, and Megere H. Community support for the HIV and AIDS continuum of response in Uganda. University Research Co; 2014. https://www.usaidassist.org/sites/assist/files/uganda_community_support_hiv_continuum_usltr_aug2014.pdf.

  14. Kalyango JN, Rutebemberwa E, Alfven T, Ssali S, Peterson S, Karamagi C. Performance of community health workers under integrated community case management of childhood illnesses in eastern Uganda. Malar J. 2012;11:282.

    Article  PubMed  PubMed Central  Google Scholar 

  15. Taylor H. Situation analysis: village health teams Uganda 2009: Health UMo; 2009.

  16. Vujicic M, Addai E, Bosomprah S. Measuring health workforce productivity: application of a simple methodology in Ghana. In: Health, nutrition and population discussion paper. Washington: Bank W; 2009.

    Google Scholar 

  17. Schellenberg JA, Victora CG, Mushi A, de Savigny D, Schellenberg D, Mshinda H, Bryce J. Tanzania integrated Management of Childhood Illness MCEBHSSG: inequities among the very poor: health care for children in rural southern Tanzania. Lancet. 2003;361(9357):561–6.

    Article  PubMed  Google Scholar 

  18. Vyas S, Kumaranayake L. Constructing socio-economic status indices: how to use principal components analysis. Health Policy Plan. 2006;21(6):459–68.

    Article  PubMed  Google Scholar 

  19. Tibshirani R. Regression shrinkage and selection via the Lasso. J Royal Stat Soc. 1996;58(1):267–88.

    Google Scholar 

  20. Goggin K, Wexler C, Nazir N, Staggs VS, Gautney B, Okoth V, Khamadi SA, Ruff A, Sweat M, Cheng AL, et al. Predictors of infant age at enrollment in early infant diagnosis Services in Kenya. AIDS Behav. 2016;

  21. van Lettow M, Bedell R, Landes M, Gawa L, Gatto S, Mayuni I, Chan AK, Tenthani L, Schouten E. Uptake and outcomes of a prevention-of mother-to-child transmission (PMTCT) program in Zomba district, Malawi. BMC Pub Health. 2011;11:426.

    Article  Google Scholar 

  22. Dieleman M, Gerretsen B, van der Wilt GJ. Human resource management interventions to improve health workers' performance in low and middle income countries: a realist review. Health Res Policy Syst. 2009;7:7.

    Article  PubMed  PubMed Central  Google Scholar 

  23. Kok MC, Dieleman M, Taegtmeyer M, Broerse JE, Kane SS, Ormel H, Tijm MM, de Koning KA. Which intervention design factors influence performance of community health workers in low- and middle-income countries? A systematic review. Health Policy Plan. 2015;30(9):1207–27.

    Article  PubMed  Google Scholar 

  24. Sharma R, Webster P, Bhattacharyya S. Factors affecting the performance of community health workers in India: a multi-stakeholder perspective. Glob Health Action. 2014;7:25352.

    Article  PubMed  Google Scholar 

  25. Crispin N, Wamae A, Ndirangu M, Wamalwa D, Wangalwa G, Watako P, Mbiti E. Effects of selected socio-demographic characteristics of community health workers on performance of home visits during pregnancy: a cross-sectional study in Busia District, Kenya. Glob J Health Sci. 2012;4(5):78–90.

    Article  PubMed  PubMed Central  Google Scholar 

  26. Kozuki N, Oseni L, Mtimuni A, Sethi R, Rashidi T, Kachale F, Rawlins B, Gupta S. Health facility service availability and readiness for intrapartum and immediate postpartum care in Malawi: a cross-sectional survey. PLoS One. 2017;12(3):e0172492.

    Article  PubMed  PubMed Central  Google Scholar 

  27. Rowe SY, Olewe MA, Kleinbaum DG, McGowan JE Jr, McFarland DA, Rochat R, Deming MS. The influence of observation and setting on community health workers' practices. Int J Qual Health Care. 2006;18(4):299–305.

    Article  CAS  PubMed  Google Scholar 

  28. Haq Z, Iqbal Z, Rahman A. Job stress among community health workers: a multi-method study from Pakistan. Int J Ment Health Syst. 2008;2(1):15.

    Article  PubMed  PubMed Central  Google Scholar 

  29. Kok MC, Muula AS. Motivation and job satisfaction of health surveillance assistants in Mwanza, Malawi: an explorative study. Malawi Med J. 2013;25(1):5–11.

    CAS  PubMed  PubMed Central  Google Scholar 

  30. Greenspan JA, McMahon SA, Chebet JJ, Mpunga M, Urassa DP, Winch PJ. Sources of community health worker motivation: a qualitative study in Morogoro region, Tanzania. Hum Resour Health. 2013;11:52.

    Article  PubMed  PubMed Central  Google Scholar 

  31. Kasteng F, Settumba S, Kallander K, Vassall A, in SSG. Valuing the work of unpaid community health workers and exploring the incentives to volunteering in rural Africa. Health Policy Plan. 2016;31(2):205–16.

    Article  PubMed  Google Scholar 

  32. Uganda Ministry of Health. Community Health Extension Wrokers Streategy in Uganda (2015/16-2019/20). 2016. https://www.medbox.org/community-health-extension-workers-strategy-in-uganda-201516-201920/preview?q=Community+Health+Extension+Workers+Strategy+in+Uganda.

  33. Bohren MA, Hunter EC, Munthe-Kaas HM, Souza JP, Vogel JP, Gulmezoglu AM. Facilitators and barriers to facility-based delivery in low- and middle-income countries: a qualitative evidence synthesis. Reprod Health. 2014;11(1):71.

    Article  PubMed  PubMed Central  Google Scholar 

  34. Kozuki N, Guenther T, Vaz L, Moran A, Soofi SB, Kayemba CN, Peterson SS, Bhutta ZA, Khanal S, Tielsch JM, et al. A systematic review of community-to-facility neonatal referral completion rates in Africa and Asia. BMC Public Health. 2015;15:989.

    Article  PubMed  PubMed Central  Google Scholar 

  35. Mangham-Jefferies L, Pitt C, Cousens S, Mills A, Schellenberg J. Cost-effectiveness of strategies to improve the utilization and provision of maternal and newborn health care in low-income and lower-middle-income countries: a systematic review. BMC Pregnancy Childbirth. 2014;14:243.

    Article  PubMed  PubMed Central  Google Scholar 

  36. Crigler L, Hill K, Furth R, Bjerregaard D. Community health worker assessment and improvement matrix (CHW AIM): USAID; 2013.

Download references

Acknowledgements

We thank Dr. John Odaga for his contribution to the study design and for overseeing data collection. TW received feedback on the study protocol from the international technical review panel and key stakeholders in Uganda: Lam Mandeya, District Health Educator, Busia Local Government, Uganda; Tony Bwire, the District Health Inspector, Busia District, Uganda; Gilbert Muyambi, World Vision Uganda; Polly Walker, World Vision International; Jackie Calnan, USAID Uganda; Diana Frymus, USAID Washington; Edward Broughton, USAID ASSIST, URC; Ram Shrestha, formerly of USAID ASSIST, URC; Allison Foster, formerly of USAID ASSIST; Wanda Jaskiewicz, formerly of IntraHealth; Charles Okira, International Health Sciences University, Kampala, Uganda; and Deo Sekimpi, Uganda National Association of Community and Occupational Health, Kampala, Uganda.

We thank Dr. Samantha Rowe, U.S. Centers for Disease Control and Prevention, Diana Frymus, Adrienne Hurst, Shayanne Martin, Lani Marquez, M. Rashad Massoud, and Jim Heiby for input into the study. We also thank World Vision International and World Vision Uganda for their contributions to the study’s technical review panel and for their logistical support during the data collection in Busia District.

Funding

Funding for this study was provided by the U.S. President’s Emergency Plan for AIDS Relief (PEPFAR) through the USAID Health Care Improvement Project and USAID Applying Science to Strengthen and Improve Systems (ASSIST) Project, both implemented by University Research Co., LLC (URC) through Contract Number GHN-I-03-07-00003-00 and Cooperative Agreement Number AID-OAA-A-12-00101, respectively. The donors had no direct involvement on the study.

Availability of data and materials

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Author information

Authors and Affiliations

Authors

Contributions

TW designed the study, the data collection instruments, oversaw data collection, prepared the supplemental materials, and reviewed the manuscript. NK conducted the data analysis and wrote the manuscript. Both authors have read and approved the manuscript.

Corresponding author

Correspondence to Naoko Kozuki.

Ethics declarations

Ethics approval and consent to participate

Ethical approval was received from University Research Co. Institutional Review Board and Makerere University Institutional Review Board. All participants were verbally consented to participate in the study.

Competing interests

The authors declare that they have no competing interests.

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Additional files

Additional file 1:

Supplemental file. Contains Tables S1, S2, S3, S4 and Supplemental Text 1. (DOCX 71 kb)

Additional file 2:

Data collection file. Contains data collection tools. (DOCX 153 kb)

Rights and permissions

Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Kozuki, N., Wuliji, T. Measuring productivity and its relationship to community health worker performance in Uganda: a cross-sectional study. BMC Health Serv Res 18, 340 (2018). https://0-doi-org.brum.beds.ac.uk/10.1186/s12913-018-3131-9

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://0-doi-org.brum.beds.ac.uk/10.1186/s12913-018-3131-9

Keywords