Abstract
Provider-patient communication (PPC) skills are key in promoting patient satisfaction. Our study examined the relationship between clinician PPC skills and patient satisfaction with care among virally unsuppressed adult HIV patients in Busia County, Kenya. This cross-sectional study was conducted among 360 HIV patients on first line antiretroviral regimen and having a recent viral load ≥400 copies HIV RNA/ml. We conducted logistic regression analysis. The mean age of participants was 48.2 years [standard deviation (SD): 12.05]. Overall, the mean score on clinician PPC skills was 33.3 (SD: 9.0). A high proportion (85%) of participants reported satisfaction with the HIV care services. After adjusting for covariates, the odds of being satisfied with care increased by 19% (adjusted odds ratio: 1.19, 95% CI: 1.11-1.30) for every one unit increase in the clinician PPC skills score. Promoting good PPC skills may be key to improving patient satisfaction with HIV care.
Introduction
How HIV care is delivered in sub-Saharan Africa (SSA) plays a critical role in patient adherence to and retention in care. Health system factors such as poor staff attitudes, congested clinics, long waiting times, discrimination and stigma, and distance to health facilities have been reported to affect adherence and retention of patients in care.1–3 Low capital and human resources, high demand for HIV services and a paternalistic approach to care have contributed to system level barriers in the region.4,5 In light of these challenges, there is need to improve the quality of HIV care services in order to enhance patient adherence and retention in care.
Patients who are the main beneficiary of HIV care systems 6 have their own expectations of health services. 7 This influences their level of satisfaction with the services they receive. 8 Patient satisfaction is a measure of how patients experience and perceive the quality of care received.6,8,9 There is evidence that patients who are satisfied with care are more likely to adhere to the treatment guidelines and hence be retained.10–12 Given the benefits achieved when patients are satisfied with care, health systems will greatly benefit from continued assessment of this construct.6,9 Feedback from these assessments may be used to inform the delivery of HIV care that ultimately promotes patient retention.
While there are a number of system-level factors that have been reported to be associated with patient satisfaction with care, including the physical status of the facility, availability of medication and time spent at the facility,11,13–19 patient-provider relationship dynamics may be the most critical to understanding patient satisfaction. Provider-patient communication (PPC) plays a critical role in facilitating better provider-patient relationship dynamics, empowers patients with knowledge about their health and promotes partnership between patients and providers in decision-making.20,21 Evidence from high income countries suggest that patients who perceive their clinician as having good communication skills also report satisfaction with care.11,13,15,17–19,22 This relationship has not yet been fully characterized in resource-limited settings with high HIV burden.11,13 We therefore examined the relationship between PPC skills and satisfaction with HIV care among virally unsuppressed adult HIV patients in Busia county, Kenya.
Methods
Study Site
The study was conducted in two rural health facilities offering HIV care in Busia County, Kenya. The two HIV health facilities, Khunyangu health facility and Port-Victoria health facility, are affiliated with the Academic Model Providing Access to Healthcare (AMPATH) program. 23 These facilities were comparable in terms of the clinic capacity with a provider-patient ratio of 1:210 per month. In addition, they each had 5 clinic officers (equivalent to physician assistants in the US) providing day to day clinical care to patients living with HIV.
Study Design
We conducted a cross-sectional study as part of the baseline assessment of a pilot randomized control trial that aimed to promote patient engagement through an enhanced patient care intervention. The baseline assessments were done between August-December 2019. A detailed description of the intervention is provided elsewhere. 24 We obtained approval from Institutional Research and Ethics Committee (approval #0003297) at Moi University in Kenya. All participants provided written informed consent prior to enrollment in this study.
Study Population and Sample
We targeted adult patients living with HIV who were receiving HIV care at either of the two health facilities. Patients were considered eligible for the study if they were on a first line ART regimen and had a first elevated (≥400 copies HIV RNA/ml) viral load (VL) within three months prior to the study rollout. A sample of 360 patients with unsuppressed viral loads was the target sample size for the pilot randomized control trial.
Data Collection Instruments
We developed a structured interviewer-administered survey that incorporated adapted validated measures of provider-patient communication (PPC) skills 25 and one item that assessed patient satisfaction with care. The item assessing patient satisfaction, “How satisfied were you with the clinical service today?” was measured using a 3-point Likert scale (1. very satisfied/satisfied, 2. neither satisfied nor dissatisfied, 3. dissatisfied/very dissatisfied, 4. don't know). PPC skills 25 measured perceptions about clinicians’ communication skills during a clinical encounter with 16 items with responses on a 3-point Likert scale (lowest score = 16, highest score = 48). The PPC scale was previously validated among adult HIV patients in Kenya and showed acceptable validity (construct and content) and reliability. Internal consistency (Cronbach's alpha = 0.81) and one-week test-retest reliability (Pearson Correlation = 0.82) were both acceptable. 25
In addition, we assessed the following domains: socio-demographic characteristics, knowledge about HIV and HIV treatment, and medication self-efficacy (HIV Self-Efficacy). 26 The selection of these constructs was informed by our previous qualitative work7,27 and were considered potential confounders of the relationship between PPC skills and patient satisfaction with care. Socio-demographic characteristics included age in years, sex (male vs female), years in HIV care and disclosed HIV status to anyone (yes vs no). Correct knowledge about HIV and HIV treatment included: correct knowledge of VL (yes vs no), correct definition of VL (yes vs no), correct frequency of VL tests based on their care plan (yes vs no), correct knowledge of their most recent VL count (yes vs no), and correct antiretroviral (ARV) timing according to their prescription (yes vs no). Medication self-efficacy 28 measured individual's efficacy to adhere to their medication, including 7 items with responses on a 3-point Likert scale (highest score = 21).
This survey instrument was first developed in English and then translated to Swahili. The survey was piloted among a predetermined number of n = 10 patients with unsuppressed viral load, who were not part of the main study sample. Following the pilot, revisions were made to clarify any items that participants had difficulty answering or were unclear. These changes mainly focused on the language (following the translation question from English to Swahili) and flow of questions in order to enhance understanding.
Recruitment and Data Collection
Using the AMPATH care data system, outreach workers (HIV positive peers) were issued with a list of patients who meet the eligibility criteria by the clinical team. The outreach workers contacted eligible patients to inform them about the study and invite them to participate. Patients who accepted the invitation to take part in the study were directed to a trained research assistant who provided detailed information about the study at the end of their clinic appointment. Only those who provided written informed consent were enrolled. Thereafter, the survey was administered in either English or Swahili, to consenting participants by trained research assistants in a private room at the respective clinics. Transport reimbursement of KES 500/USD 5 was given to patients who participated in the study.
Data Analysis
Descriptive statistics such as frequencies and the corresponding percentages were used to summarize categorical variables describing the study sample, while median and interquartile range (IQR) were used to summarize continuous variables.
Clinician PPC skills was considered as a continuous variable. The internal consistency of this scale was high (Cronbach's alpha = 0.89). Patient satisfaction with HIV care was transformed into a binary variable: satisfied (very satisfied/satisfied) versus dissatisfied (neither satisfied nor dissatisfied and dissatisfied/very dissatisfied). For medication efficacy (Cronbach's alpha = 0.63) we transformed the scores into a binary variable using cut-points defined by the distribution of scores. Scores in the highest quartile of medication efficacy were considered as high medication self-efficacy.
Logistic regression analysis was used to assess the association between clinician PPC skills (main exposure) and patient satisfaction (outcome variable). We first explored bivariate associations between our independent and outcome variable. Thereafter we examined the association between clinician PPC skills and patient satisfaction with care, accounting for potential confounders. We adjusted for age in years, sex (male vs female), marital status (single/divorced/widowed vs married/partnered), level of education (primary education or less vs secondary education or more), correct knowledge of most recent VL count (no vs yes), and medication self-efficacy (yes vs no). We excluded variables with limited variability, multicollinearity and a high proportion of missingness (>10%) from adjusted analysis. All statistical tests were conducted at p ≤ 0.05 level of significance using R (version 3.6.3).
Results
Study Sample Characteristics
Of the 360 eligible patients contacted, 328 (91.1%) were recruited and participated in the study. Table 1 shows the baseline characteristics of the study participants. Overall, the mean age of participants was 48.2 years (Standard deviation [SD]: 12.1). The study sample was predominantly female (56%) and married/partnered (60%) and receiving care from study site 1 (69%). In addition, those having at least a primary level of education (82%), walking 30 min and more to their HIV facility (59%) and having disclosed their HIV status (99%) were the majority. In terms of knowledge of their HIV care, slightly more than half reported to have knowledge of what a viral load (VL) was, and among those, the majority (74%) provided the correct definition of a VL. About half (54%) of the participants correctly stated how often their VL test was taken and 59% had the correct knowledge of their most recent viral load count. Almost all (99%) of the participants provided the correct timing of when they were expected to take their ARV medication; however, none could name the ARV medication they were taking.
Patient Sociodemographic and HIV Care Knowledge Characteristics among Unsuppressed Patients.
Relationship Between Clinician PPC Skills and Patient Satisfaction with HIV Care
Generally, a high proportion 278 (85%) of participants reported to be satisfied with the HIV services they received. The mean score on clinician PPC skills was 33.3 [standard deviation (SD) = 9.0].
Participants who were satisfied with care reported higher clinician PPC skills (mean score = 35.02 [SD = 8.22]), compared to those who were not satisfied with care (mean = 23.44; SD = 7.04) (Figure 1).

Clinician PPC skills mean scores by patient satisfaction with HIV care services among unsuppressed HIV patients.
Table 2 shows that in the unadjusted analysis, for every one unit increase in clinician PPC skills score, the odds of being satisfied with care increased by 19% (OR = 1.19, 95% CI: 1.14-1.26). This relationship was the same after adjusting for age, marital status, education level, years in care, correct knowledge of the most recent VL count, and medication self-efficacy [Adjusted odds ratio (AOR) = 1.19; 95% CI, 1.11 to 1.30].
Unadjusted and Adjusted Odds Ratios of Patient Satisfaction with Care and Clinician PPC Skills, and Other Characteristics among Unsuppressed HIV Patients.
Unadjusted Odds Ratio (UOR)
Adjusted Odds Ratio (AOR)
Discussion
Our study revealed that patients’ assessments of clinicians’ PPC skills was positively associated with patient satisfaction with care among adult HIV patients with unsuppressed VL. We also observed that a high proportion of participants reported satisfaction with their HIV care. These findings are critical for informing HIV care programs on the importance of PPC skills in improving the quality of service delivery and overall patient satisfaction with care.6,9
Similar to our findings, other studies have found a direct positive relationship between PPC skills and patient satisfaction.11,13,15,17–19,22 Patients who reported that their clinicians had better communication skills also reported to be satisfied with care. This highlights the importance of PPC skills as the core in clinical practice and a key component of patient-centered care that promotes patient satisfaction with care.4,20,21 Following the benefits reported among HIV patients satisfied with care including ART adherence10–12 there is need for HIV care programs in this region to promote better interaction between providers and their patients through enhanced PPC skills.
The overall clinician PPC skills mean score was lower than what had been previously reported among HIV patients in the same region, 29 which is likely explained by the design of the trial from which our data are taken. Participants in the trial all had unsuppressed VL and hence the views of the virally suppressed patients may not be reflected in our data. Meanwhile, amidst the financial, human resource and structural challenges health systems in the region face,4,5 on-the-job trainings could be one of the approaches recommended to ensure that clinicians improve on their PPC skills. 30 In addition, the use of standard patient actors, successfully piloted among providers offering HIV care to young adults in Kenya, could be adopted to improve clinicians’ PPC skills. 31 This will not only promote better patient satisfaction with care, but ultimately improve on patient health outcomes. 22
Similar to other studies across the region,11,13,14,32 we found that a majority of our participants were satisfied with the HIV care they were receiving at their respective health facilities. It is however important to note that satisfaction is greatly influenced by socioeconomic status. 13 A study in Nigeria noted that individuals of a high economic status were less likely to attend public hospital as well as be satisfied with care. 13 Thus patients in public facilities were more satisfied with HIV care compared to those in private facilities. 13 Majority of the participants in our study had a less than secondary level of education and were receiving HIV care free of charge at rural public health facilities. This may reflect on their low socioeconomic status. Given its impact on patient health outcomes,10–12 additional studies with a wider category of HIV patients may also be needed to provide a holistic perspective patient satisfaction and its association with PPC skills in this population.
Finally, it is important to note that none of our study participants could accurately name the ARV medications they were receiving. Studies in developed countries have also documented inaccurate reporting of ARV medication among patients living with HIV. 33 In our context, this could be partly due to the low level of education among study participants and the complex names assigned to ARV medications. Given that patient ARV knowledge is key to medication adherence, 33 there may be need for HIV care programs to identity simpler ways for patients to correctly recall their medication. Efforts to promote and maintain patients’ active involvement in HIV care should also be continuously encouraged.
There are several study limitations. Generally, patient satisfaction with care as a construct may be difficult to measure since it varies over time and circumstances. 8 In addition, the surveys were conducted in the health facilities where participants received their routine HIV care which may have introduced social desirability bias. Furthermore, our study design and eligibility criteria limited our analysis, hence we could not explore the relationship between patient satisfaction and treatment outcomes. We also did not account for clinicians’ experiences and communication skills that may have influenced the PPC dynamics. Finally, although we sampled HIV patients receiving care from public HIV clinics, the clinics we selected were part of the AMPATH program. The way AMPATH provides care may differ from other HIV care programs in Kenya hence our results may not be generalizable to other rural settings in Kenya. Despite these challenges our study adds to the body of literature on the value of patient satisfaction in HIV care and the important factors to consider.
Conclusion
In conclusion, our study showed that clinician PPC skills were associated with patient satisfaction with care among virally unsuppressed adult HIV patients. Promoting better provider-patient interactions may be key to improving patients’ level of satisfaction with the HIV services offered. Training of providers on effective PPC skills is an initial step towards achieving this goal. Additional research is needed to better understand approaches to improve PPC skills in an effort to enhance the experiences patients have with their HIV care facilities.
Footnotes
Authors’ Contributions
All authors aided in the conceptualization of the manuscript. JW developed the initial drafts. JW, AM, BG and IW reviewed and edited the manuscript for publication. All authors reviewed the final drafts and approved the manuscript for submission.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the National Institute of Health (grant number K43TW010684, U54GM115677, P30AI042853).
