Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2019 Sep 26.
Published in final edited form as: Int J Ment Health Addict. 2017 May 24;15(4):766–781.

FACTORS INFLUENCING LAY AND PROFESSIONAL HEALTH WORKERS’ SELF-EFFICACY IN IDENTIFICATION AND INTERVENTION FOR ALCOHOL, TOBACCO, AND OTHER SUBSTANCE USE DISORDERS IN KENYA

Sian Hsiang-Te Tsuei 1, Veronic Clair 1,2, Victoria Mutiso 2, Abednego Musau 2, Albert Tele 2, Erica Frank 1, David Ndetei 2,3
PMCID: PMC6761831  NIHMSID: NIHMS941883  PMID: 31558889

Abstract

The global burden of substance use disorders (SUDs), including alcohol and tobacco, disproportionately affect low- and middle-income countries (LMICs), considering their rising disease burden and low service capacity. Nested within a Kenyan training program, this study explores factors associated with healthcare providers’ self-efficacy to treat SUD. Surveys of 206 healthcare workers were used to perform regression and sensitivity analysis assessing various factors association with self-efficacy. Self-efficacy for SUD was lower in those practicing in public facilities and perceiving a need for alcohol use disorder (AUD) training; while higher self-efficacy correlated with a higher proportion of patients with AUD in one’s setting, access to mental health worker support, cannabis use at a moderate risk level, and belief that AUD is manageable in outpatient settings. Increasing awareness about SUD prevalence, identification, and treatment skills could improve the self-efficacy of LMICs’ health care providers and therefore the willingness to implement more services for patients with SUDs.

Keywords: Alcohol, Tobacco, Substance use disorders, Self-efficacy, Low-middle-income countries

The global burden of substance use disorders (SUDs) increased by 40% from 1990 to 2010, with tobacco and alcohol use, respectively, now ranked as the 2nd and 5th most important risk factors, while all other SUDs (OSUDs) ranked 19th (Lim et al. 2012). Despite the World Health Organization’s (WHO) recommendation for primary care workers to leverage their uniquely longitudinal, credible relationships to minimize patient harm from tobacco, alcohol, and OSUDs (Henry-Edwards et al. 2003), more than 75% of patients with SUDs in low- and middle-income countries (LMICs) have no access to treatment (WHO 2010). Kenya has no official SUD-related guidelines for primary care despite some risk of SUD-related problems: the rates of current use for tobacco, alcohol, khat,1 and cannabis in 15–65 year olds are 8.6, 13.6, 4.2, and 1.1%, respectively; lifetime use rates for the same substances are approximately 16.7, 29.9, 3.5, and 6.0%, respectively (National Campaign Against Drug Abuse Authority 2012; Othieno et al. 2000).

In addition to the barriers of stigma and insufficient health capacity in SUDs and mental health in general (Saxena et al. 2007; The Academy of Medical, 2008; WHO 2011), practitioners’ self-efficacy also affects how actively and effectively practitioners perform their duties (Delgadillo et al. 2014; Elder et al. 1999). Knowledge and skills are not enough to change behaviors. Self-efficacy, an integral concept of behavior change theory, has been studied in several field as a predictor of someone’s ability to change their behaviors independently of other factors (Battersby et al. 2010; Bazzano et al. 2009; Beckman et al. 2006; DiClemente et al. 1995; Kadden and Litt 2011). Studies have shown that the concept of self-efficacy applies to physicians and other health professionals in relation to improving their practice (Grol and Wensing 2016; Lemieux et al. 2011; O’Campo et al. 2011), as well as use of counseling for patients on a variety of topics (Bahora et al. 2008; Buckelew et al. 2008; Miller Perrin et al. 2005; Ozer et al. 2004), including SUDs (Buckelew et al. 2008; Gottlieb et al. 1987; Harris et al. 2016; Ozer et al. 2004; Schram et al. 2015; Thompson et al. 1993; Woddruff et al. 2010). Because insufficient self-efficacy compromises health workers’ performance, this project (the Computer-based Drug and Alcohol Training Assessment in Kenya [eDATA K]) aims to provide health care workers with more training opportunities to jumpstart a virtuous cycle that simultaneously strengthens confidence, skills, and desire to provide more comprehensive and effective SUDs screening (Delgadillo et al. 2014). This is expected to drive sustainable long-term efforts to identify patients affected by SUDs and provide them with evidence-based preventive and clinical interventions (Bandura and Cervone 1983; Kay and Shipman 2014).

Nested within this overall project is this current study that aims to establish a baseline level of self-efficacy in primary care workers (PCWs) regarding the provision of screening, brief intervention, and care to people with SUDs, and to close the knowledge gaps around driving factors behind PCW’s self-efficacy. Elucidating this relationship can reveal more effective motivations for practitioners to intervene more frequently and effectively for at-risk individuals, and to facilitate the strengthening of preventive and clinical care that lowers the overall burden of disease in resource-starved countries.

METHODS

Participants

This study was approved by the Kenya Medical Research Institute Ethics Review Committee (KEMRI-ERC) and the University of British Columbia (UBC) Ethics Board. Administrative support was provided by the various district health management teams, county government health officials, and facility managers. The study was carried out in 2014 as a sub-study of the eDATA K project in the 15 eDATA K facilities selected based on (1) offering typical primary care services, (2) PCWs expressing interest and capacity to engage in the training program, (3) having electricity, and (4) representing a mix of public and private facilities of the participating geographic areas. Of the facilities, 11 were public primary care outpatient clinics in Machakos and Makueni Counties and 4 were private outpatient clinics in Nairobi and Machakos.

All PCWs (n = 236) participating in eDATA K in the selected facilities’ outpatient services were invited to participate in the self-efficacy survey. Participants provided informed verbal or written consent before data collection, and questionnaires were serialized and then distributed to respondents. Participants were given half a day to respond to the questionnaire to minimize interference with routine work related activities, and to ensure good response rates. PCWs participating in eDATA K represented >80% of the staff in public health centers and dispensaries, while in larger facilities (e.g., hospitals’ outpatient departments), 50–75% of outpatient staff participated.

Measures

WHO, through the mental health Global Action Program (mhGAP), offers a set of SUD care principles, services, and algorithms recommended for primary care workers in LMICs (WHO 2008). The development of the survey questions was based on self-efficacy (Bandura and Cervone 1983; Franco et al. 2002) theory in performing the SUD services in the mhGAP. The self-efficacy survey items had been pilot tested in a prior phase of the study and answered on a 3-point Likert scale: (1) I am unsure I can do it, (2) I am confident I can do it, and (3) I am very confident I can do it. The 11 survey items selected for the creation of the self-efficacy score for this study were improved based on that pilot study, had excellent face validity when reviewed by experts and practitioners, with the three-level Likert scale being adequate to avoid ceiling effects, and demonstrated excellent reliability with an overall Cronbach alpha of 0.943. A composite score, out of 10, was created if respondents answered at least 7 questions.

Independent variables measured included demographic variables (gender, age, religion, socioeconomic status, lifetime use of substances, types of facility). The socioeconomic status variable was a 10-point score constructed by aggregating assets, a common method of assessing economic status in LMIC (Howe et al. 2012). Each asset in the survey was weighted in reverse function of proportion of population owning it (1, 2, 3, and 4 points were given for owning a mobile phone, bicycle, motorbike, and a car, respectively (Howe et al. 2012). Substance-related measures included perceived prevalence of SUDs in patients presenting to their facility, perceived responsibilities toward treating those with SUDs, and perceived training needs for SUDs. The WHO Alcohol Smoking and Substance Involvement Screening Test (ASSIST) (Humeniuk et al. 2010) was administered to all health workers. ASSIST risk scores derived from self-reported usage patterns (quantity, frequency, craving, prior unsuccessful attempt to quit, experiencing negative health, social or other consequences) of tobacco, alcohol, and cannabis were calculated and stratified into low-, moderate-, or high-risk categories (Humeniuk et al. 2008, 2010). Questions related to respondents’ perception of SUD prevalence and HCW responsibility for SUD treatment and other knowledge, attitude or practice (KAP) questions were adapted from prior studies in Kenya that targeted KAP related to mental illnesses and that had been used in the UK in the past (Morgan and Killoughery 2003; Ndetei et al. 2011).

Categorical variables with more than two levels were converted to dummy variables. Doctors, nurses, and clinical officers were classified as clinicians and all other staff (mostly community health workers in community health centers and dispensaries, and mostly receptionists in private facilities and outpatient departments of hospitals) were classified as support staff. Marital status was categorized as either (a) single, widowed, or divorced vs. (b) cohabiting or married.

Statistical Analysis

Analyses were conducted using IBM SPSS® Version 21. Descriptive statistics were used to examine the general data distribution, and Pearson’s and Spearman’s correlations were used to explore associations between independent and dependent variables for continuous and categorical variables, respectively. Support staff’s and clinicians’ levels of self-efficacy for each items were compared using independent t test while paired t test was used to compare each item’s mean to the overall 11-item mean (computed for all respondents who answered at least 7 items). Three multivariate regression models were created using self-efficacy scores (out of 10) as the dependent variable. The first model included all demographic variables and significantly correlated variables (p ≤ 0.05). The second model included only significantly correlated explanatory variables from the bivariate analysis. The third regression aimed to generate the most parsimonious model using a stepwise approach, where variables with F ≤ 0.05 level of significance were entered and F ≥ 0.10 dropped. Within the stepwise regression process, theory-based possible interactions were examined. Interaction terms tested were created by matching the first set of variables (educational status, facility) with the second set of variables (opportunity for refresher course, need for training, perceived need to manage substance use disorders as part of the work). The interaction terms tested are summarized by the relationships in Fig. 1. The model was built using the mean substitution method for missing values.

Figure 1.

Figure 1

Open in a new tab

Interactions terms tested

RESULTS

Among eDATA K participants, 87.3% (n = 206) completed the self-efficacy survey. Five respondents answered <7 self-efficacy items and were considered missing. The study population’s demographics, baseline SUD-related training and capacity, and these characteristics’ correlation with self-efficacy are reported in Table 1. Most respondents were female, had a mean age of 35 years old, held a certificate level of education (mostly nurses by profession), lived with a spouse, and worked in public facilities or hospital outpatient clinics. The only demographic significantly correlated with self-efficacy was a negative correlation with being a public facility employee.

Table 1.

Characteristics of respondents and bivariate correlations

Participants’ characteristicsa Support staff Clinicians Total SEb mean ± SE Rc p value
Gender (n, %) Male 43, 39.8% 31, 34.2% 74, 37.2% 6.30 ± 0.21 0.007 0.924
Female 65, 60.2% 60, 65.9% 125, 62.8% 6.40 ± 0.16
Education level (n, %) Secondary school 32, 29.6% 0, 0.0% 32, 16.0% 6.17 ± 0.30 −0.036 0.620
Certificate 60, 55.6% 30, 32.6% 90, 45.0% 6.53 ± 0.19
Diploma 5, 4.7% 33, 35.9% 38, 19.0% 6.48 ± 0.29
Degree 11, 10.2% 29, 31.5% 40, 20.0% 6.02 ± 0.30
Religious affiliations (n, %) Christian 109, 98.2% 91, 100.0% 200, 99.0% 6.34 ± 0.12 0.008 0.911
Muslim 2, 1.8% 0, 0.0% 2, 1.0% 6.21 ± 0.45
Marital status (n, %) Single, divorced, widow 41, 36.3% 24, 26.1% 65, 31.7% 6.64 ± 0.22 −0.105 0.140
Living together married, living together 72, 63.7% 68, 73.9% 140, 68.3% 6.18 ± 0.15
Facility type (n, %) Private 28, 24.8% 22, 23.6% 50, 24.3% 7.09 ± 0.24 −0.237 0.001**
Public 85, 75.2% 71, 76.3% 156, 75.7% 6.11 ± 0.14
Facility level (n, %) Community clinic 61, 54.0% 35, 37.6% 96, 46.6% 6.27 ± 0.17 0.030 0.673
Hospital outpatient clinic 52, 46.0% 58, 62.4% 110, 53.4% 6.36 ± 0.17
Age (mean ± SE) 34.17 ± 0.992 36.64 ± 1.092 35.30 ± 0.738 – 0.022 0.772
SES10 (mean ± SE) 1.70 ± 0.187 2.14 ± 0.217 1.91 ± 0.143 – −0.137 0.069
Health workers’ substance use risk (n, %)
Alcohol Low 110, 97.3% 88, 96.7% 198, 97.1% 6.32 ± 0.12 −0.003 0.971
Moderate 3, 2.7% 3, 3.3% 6, 3.0% 6.48 ± 0.62
Tobaccod Low 102, 90.3% 77, 84.6% 179, 87.7% 6.30 ± 0.13 0.023 0.747
Moderate 10, 8.8% 14, 15.4% 24, 11.8% 6.54 ± 0.29
High 1, 0.9% 0, 0.0% 1, 0.5% –
Cannabis Low 109, 96.5% 88, 96.7% 197, 96.6% 6.26 ± 0.12 0.155 0.029*
Moderate 4, 3.5% 3, 3.3% 7, 3.4% 7.99 ± 0.78
TUD in practice/training
% of patients seen (mean ± SE) Predominantly TUD cases 26.33 ± 2.39 22.37 ± 2.55 24.45 ± 1.75 – 0.264 <0.001**
Significant TUD and physical condition 21.45 ± 1.94 18.07 ± 2.06 19.84 ± 1.42 – 0.174 0.023*
Mild TUD and physical condition no TUD 22.50 ± 2.42 18.90 ± 2.33 20.76 ± 1.68 – 0.103 0.179
47.53 ± 3.30 48.88 ± 3.91 48.18 ± 2.53 – −0.041 0.599
Responsible for TUD assessment in new patients (n, %) Yes 70, 67.3% 78, 89.7% 155, 81.1% 6.41 ± 0.14 −0.149 0.042*
No 34, 32.7% 9, 10.3% 36, 18.8% 5.87 ± 0.27
Comfort level in managing TUD (mean ± SE) 2.46 ± 0.095 2.80 ± 0.096 2.63 ± 0.07 – −0.465 <0.001**
Management of TUD is an important part of my care for chronic outpatients (n, %) Yes 76, 72.4% 82, 91.1% 158, 81.0% 6.47 ± 0.14 −0.225 0.002**
No 29, 27.6% 8, 8.9% 37, 19.0% 5.63 ± 0.27
Access to a mental health worker to refer TUD cases? Yes 51, 58.6% 51, 70.8% 102, 64.2% 6.37 ± 0.18 −0.031 0.698
No 36, 41.4% 21, 29.2% 57, 35.8% 6.25 ± 0.21
TUD can be successfully managed in outpatients clinics of general hospitals Yes 66, 66.7% 64, 71.9% 130, 69.1% 6.48 ± 0.15 −0.140 0.059
No 33, 33.3% 25, 28.1% 58, 30.9% 5.99 ± 0.23
Had a chance to attend a refresher course in TUD Yes 23, 21.9% 16, 18.2% 39, 20.2% 6.43 ± 0.29 −0.037 0.611
No 82, 78.1% 72, 81.8% 154, 79.8% 6.28 ± 0.14
Need for TUD component in the training in your field Yes 80, 81.6% 75, 93.8% 155, 87.1% 6.30 ± 0.13 0.058 0.447
No 18, 18.4% 5, 6.3% 23, 12.9% 6.56 ± 0.37
AUD in practice/training % of patients seen (mean ± SE)
Predominantly AUD cases 27.63 ± 3.05 22.03 ± 2.48 25.44 ± 1.84 – −0.231 0.002**
Significant AUD and physical condition 24.03 ± 2.55 23.01 ± 2.74 23.67 ± 1.70 – 0.184 0.014*
Mild AUD and physical condition 24.77 ± 2.54 23.22 ± 2.39 24.32 ± 1.60 – 0.202 0.007**
No AUD 49.33 ± 3.05 47.16 ± 3.70 48.29 ± 2.38 – −0.100 0.194
Responsible for AUD assessment in new patients (n, %) Yes 75, 69.4% 84, 92.3% 159, 79.9% 6.38 ± 0.13 −0.060 0.405
No 33, 30.6% 7, 7.7% 40, 20.1% 6.14 ± 0.28
Comfort level in managing AUD (mean ± SE) 2.63 ± 0.089 2.84 ± 0.091 2.72 ± 0.063 – −0.395 <0.001**
Management of AUD is an important part of my care for chronic outpatients Yes 78, 75.0% 82, 90.1% 160, 82.1% 6.50 ± 0.14 0.157 0.029*
No 26, 25.0% 9, 9.9% 35, 17.9% 5.72 ± 0.28
Access to a mental health worker to refer AUD cases? Yes 55, 61.1% 59, 76.6% 114, 68.3% 6.52 ± 0.16 0.204 0.009**
No 35, 38.9% 18, 23.4% 53, 31.7% 5.82 ± 0.23
AUD can be successfully managed in outpatients clinics of general hospitals Yes 59, 55.1% 55, 61.8% 114, 58.2% 6.73 ± 0.17 0.245 0.001**
No 48, 44.9% 34, 38.2% 82, 41.8% 5.84 ± 0.17
Had a chance to attend a refresher course in AUD Yes 18, 16.2% 8, 8.8% 26, 12.9% 7.04 ± 0.32 0.183 0.010**
No 93, 83.8% 83, 91.2% 176, 87.1% 6.26 ± 0.13
Need for AUD component in the training in your field Yes 90, 84.1% 81, 95.3% 171, 89.1% 6.19 ± 0.13 −0.222 0.002**
No 17, 15.9% 4, 4.7% 21, 10.9% 7.39 ± 0.37
OSUD in practice/training % of patients seen (mean ± SE)
Predominantly OSUD cases 24.79 ± 2.30 18.99 ± 2.40 22.14 ± 1.73 – 0.347 <0.001**
Significant OSUD and physical condition 23.46 ± 2.23 18.13 ± 2.29 20.92 ± 1.65 – 0.287 <0.001**
Mild OSUD and physical condition 23.98 ± 2.12 17.74 ± 2.09 20.91 ± 1.54 – 0.180 0.018*
No OSUD 49.16 ± 2.93 50.46 ± 3.54 46.97 ± 2.32 – −0.160 0.037*
Responsible for OSUD assessment in new patients (n, %) Yes 70, 67.3% 82, 90.1% 152, 77.9% 6.45 ± 0.15 −0.150 0.038*
No 34, 32.7% 9, 9.9% 43, 22.1% 5.80 ± 0.25
Comfort level in managing OSUD (mean ± SE) 2.52 ± 0.093 2.83 ± 0.088 2.66 ± 0.070 – −0.485 <0.001*
Management of OSUD is an important part of my care for chronic outpatients (n, %) Yes 81, 77.9% 80, 89.9% 161, 83.4% 6.41 ± 0.14 −0.147 0.042*
No 23, 22.1% 9, 10.1% 32, 16.6% 5.83 ± 0.31
Access to a mental health worker to refer OSUD cases? Yes 61, 64.9% 59, 72.8% 120, 68.6% 6.43 ± 0.15 −0.089 0.246
No 33, 35.1% 22, 27.2% 55, 31.4% 6.12 ± 0.25
OSUD can be successfully managed in outpatients clinics’ of general hospitals (n, %) Yes 61, 57.5% 60, 65.9% 121, 61.4% 6.54 ± 0.17 −0.168 0.020*
No 45, 42.5% 31, 34.1% 76, 38.6% 5.96 ± 0.18
Had a chance to attend a refresher course in OSUD Yes 21, 19.6% 7, 7.7% 28, 14.1% 6.62 ± 0.39 −0.067 0.354
No 86, 80.4% 84, 92.3% 170, 85.9% 6.32 ± 0.13
Need for OSUD component in the training in your field Yes 82, 85.4% 82, 97.6% 164, 91.1% 6.28 ± 0.13 0.085 0.261
No 14, 14.6% 2, 2.4% 16, 8.9% 6.79 ± 0.46
Open in a new tab
a

Mean values were reported for variables of at least ordinal level. Proportions were reported for categorical variables

b

SE = mean self-efficacy score, can be calculated only for categorical variables

c

Correlation: Spearman coefficients were used for non-parametric variables and Pearson coefficients were was used for the parametric variables

d

High-risk level of use was found only for one tobacco user and for none of the users of the other substances

The vast majority of PCW fell into the ASSIST category of low level of health risk from using alcohol, tobacco, or cannabis. Only the risk of cannabis use significantly correlated with self-efficacy, with those at moderate risk of experiencing health consequences from their cannabis use reporting more self-efficacy. PCW reported that about a quarter of their patients had predominantly a tobacco use disorder (TUD) (24.45%) (versus significant, mild, or no TUD use), and this proportion was similar for alcohol use disorder (AUD) (25.44%) and OSUDs (22.14%). The perceived prevalence of patients seen in practice for each type of SUD, the comfort level in managing patients with a given type of SUD, and agreeing that managing a given SUD is part of one’s responsibility were significantly correlated with self-efficacy for each respective type of SUD. For AUD specifically, considering that AUD management is an important part of the care of outpatients, believing AUD can be successfully managed in outpatient settings, having access to a mental health worker for referral of AUD, having attended a refresher course, and perceiving no further personal need for AUD training in one’s field were all positively correlated with self-efficacy. Variables significant for AUD were not similarly significant when it came to TUDs and OSUDs. Notably, the questions related to having had access to refresher course, or perceiving the need for training in one’s field for these substances were not correlated with self-efficacy. However, for TUD and OSUDs, feeling responsible for the assessment of these disorders in new patients was associated with higher self-efficacy, which was not the case for AUD.

Self-Efficacy Scores

Support staff and clinicians did not differ significantly in their self-efficacy (Table 2), averaging a relatively good level of self-efficacy (1.9 out of a scale of 3), with some variability across items. Respondents reported lower self-efficacy in determining the roles and responsibilities of primary care in relation to SUDs and in performing effective brief interventions around SUDs. They reported higher self-efficacy in communicating effectively with patients with SUDs and their families, providing advice on how SUDs affects social functioning, and recognizing SUDs-related complications.

Table 2.

Individual items and average self-efficacy scores

Self-efficacy scale itemsa,b Support staff
(mean ± SE)		 p valuec Clinicians
(mean ± SE)		 Total (mean ± SE)
pd
1: Screen for alcohol, tobacco, or other SUD 1.87 ± 0.07 0.536 1.93 ± 0.07 1.89 ± 0.05
0.912
2: Determine risky levels of alcohol, tobacco and substance consumption 1.86 ± 0.07 0.656 1.90 ± 0.07 1.86 ± 0.05
0.407
3: Recognize conditions that might be co-morbidities of alcohol, tobacco or other SUDs 1.84 ± 0.07 0.304 1.94 ± 0.06 1.92 ± 0.05
0.480
4: Recognize alcohol, tobacco or other SUDs complications 1.92 ± 0.06 0.125 2.05 ± 0.06 2.01 ± 0.05
0.014*
5: Communicate effectively with people suffering from SUDs and their families 2.04 ± 0.06 0.143 1.89 ± 0.08 1.97 ± 0.05
0.024*
6: Perform effective brief intervention for SUDs 1.83 ± 0.07 0.564 1.78 ± 0.07 1.81 ± 0.05
<0.001**
7: Identify high risk SUDs complications needing immediate medical attention 1.93 ± 0.07 0.622 1.98 ± 0.06 1.96 ± 0.05
0.069
8: Describe the roles and responsibilities of primary care in relation to SUD 1.79 ± 0.06 0.409 1.71 ± 0.07 1.77 ± 0.05
<0.001**
9: Identify when it is appropriate to refer someone for SUDs 1.86 ± 0.07 0.711 1.89 ± 0.07 1.90 ± 0.05
0.284
10: Advice on how SUDs affects social functioning 2.03 ± 0.06 0.665 1.99 ± 0.07 2.02 ± 0.05
<0.001**
11: Discuss trauma in the context of substance use disorders in a culturally appropriate and compassionate way 1.98 ± 0.07 0.169 1.85 ± 0.07 1.93 ± 0.05
0.407
Average score of all 11 items 1.90 ± 0.05 1.90 ± 0.05 1.90 ± 0.04
Open in a new tab
a

Respondents were asked to rank their ability to perform the skills listed below in a 3-point Likert scale (1 = I am unsure I can do it, 2 = I am confident I can do it, 3 = I am very confident I can do it)

b

Mean and standard error of the mean (SE) are reported in the table for the full data available, for the t tests, mean were calculated with missing data removed analysis by analysis

c

p value for independent t test between the support staff and clinicians

d

p value for comparison of the items as total and the average of the PC scores for the 11 individual items

Regression Analysis

The three regressions identified the same independent variables with similar magnitude and directionality (Table 3). The final and most parsimonious model indicates that those working in public facilities and those perceiving a personal need for professional training on alcohol use in their field had lower self-efficacy by 18 and 22%, respectively, than the baseline model estimate. When care providers have access to a mental health worker, or thought that AUD can be managed in outpatient settings, their self-efficacy scores increased by 12 and 13%, respectively. Increasing by 10% the perceived prevalence of patients significantly affected by AUD was correlated with a 2% increase in self-efficacy scores. Those who had a moderate risk of cannabis use had an increase of 29% in their self-efficacy scores over baseline; the belief that managing patients with AUD is part of one’s work responsibility, and having had an opportunity to take a refresher course did not remain significantly correlated and were not retained in the final model. None of the interaction terms tested attained statistical significance.

Table 3.

Multiple regression analysis of factors associated with self-efficacy

Variables Model 1a Model 2b Model 3c
B (SE) β Sig. B (SE) β Sig. B (SE) β Sig.
(Constant) 4.282 (1.025) <0.001 4.792 (0.748) <0.001 5.387 (0.703) <0.001
Facility—public vs. private −1.011 (0.276) −0.255 <0.001 −0.886 (0.254) −0.224 0.001 −0.976 (0.252) −0.246 <0.001
SES10 −0.035 (0.063) −0.037 0.578
Gender—male vs. female 0.210 (0.237) 0.060 0.376
Education −0.145 (0.116) −0.083 0.213
Cannabis risk—moderate vs, low 1.618 (0.603) 0.173 0.008 1.541 (0.588) 0.164 0.009 1.583 (0.588) 0.169 0.008
In your own estimation, out of every 100 patients that you see, how many do you think are significantly affected by AUD 0.012 (0.005) 0.160 0.013 0.012 (0.005) 0.150 0.018 0.011 (0.005) 0.146 0.022
Do you have access to a mental health worker in case you need to refer alcohol use cases? 0.614 (0.267) 0.151 0.023 0.633 (0.259) 0.156 0.015 0.632 (0.260) 0.156 0.016
Is there a need for alcohol use component in the training of professionals in your field? −1.079 (0.378) −0.192 0.005 −1.257 (0.365) −0.223 0.001 −1.195 (0.351) −0.212 0.001
AUD can be successfully managed in outpatients clinics’ in a general hospital 0.657 (0.226) 0.186 0.004 0.644 (0.223) 0.182 0.004 0.695 (0.223) 0.197 0.002
Have you had a chance to attend a refresher course in alcohol use treatment and prevention? 0.277 (0.333) 0.054 0.407 0.299 (0.327) 0.058 0.361
Alcohol use problems are part of my work 0.322 (0.271) 0.079 0.237 0.320 (0.267) 0.079 0.233
Management of alcohol use disorders is an important part of my care for chronic outpatients 0.497 (0.297) 0.109 0.096 0.443 (0.294) 0.097 0.133
Model parameters F = 5.515, p < 0.001
Adj R2 = 0.223, R2 = 0.272		 F = 7.525, p < 0.001
Adj R2 = 0.223, R2 = 0.257		 F = 10.252, P < 0.001
Adj R2 = 0.213, R2 = 0.236
Open in a new tab

Bunstandardized coefficients, β standardized coefficients

*

p ≤ 0.05;

**

p ≤ 0.01

a

Enter all

b

Enter all with only significantly correlated variables

c

Stepwise enter, include with p < 0.05, exclude with p > 0.10

DISCUSSION

This study used self-efficacy theory (Bandura and Cervone 1983; Franco et al. 2002) in the assessment of PCW confidence in practicing several key WHO mhGAP competencies to address SUDs in primary care settings in LMICs. This area needs further attention, due to the global burden of disease from SUDs (Lim et al. 2012), and the lack of evidence around factors facilitating the implementation of SUD-related care.

Our study suggests that an increased sense of responsibility of PCW in relation to addressing TUD and OSUDs, and increased training in AUDs are correlated with increased self-efficacy, as expected from prior studies (Grol and Grimshaw 2003). It is encouraging that some prior work has shown that practitioners’ subjective sense of counseling frequency and competence are reliably correlated with objective assessments (Frank et al. 2005). However, these factors did not remain significant in our multivariate models—suggesting other factors that remained significant account for more of the correlation.

Public Facility

Public PCW had lower self-efficacy than their counterparts in private facilities. This could be due to the historic internal “brain drain” of health professionals from public to private sectors in Kenya because private health professionals have had historically more opportunities for higher wages, access to more advanced and better maintained equipment, and superior staff morale (Ndetei et al. 2008a; Schrecker and Labonte 2004). Those left in the public sphere may therefore simultaneously have fewer qualifications, less support, and a higher patient load. Furthermore, the public system seems to give little weight to tackling SUDs (Jenkins et al. 2013), marked by deficient policies to control alcohol availability (Njenga 2015), mixed messages in relation to khat (Staff of the Global Legal Research, 2015), and minimal evidence-based tobacco control (Tumwine 2011). Indeed, the most recent strategic and investment plan for Kenya’s health care cited only two criteria related to SUDs (i.e., % population who smoke, and % population consuming alcohol regularly (Republic of Kenya 2013). The document aggregated these criteria within the category of all risk factors, and set a tragic target of allowing these risk factors to rise in prevalence from 36 to 80% (Republic of Kenya 2013). It is therefore unsurprising that PCWs report the lowest level of self-efficacy in relation to describing the roles and responsibilities of primary care in relation to SUDs.

Substance Use Influence on Self-Efficacy

Increased familiarity with the bio-psycho-social experience of substance users could lead to decreased stigma and increased self-efficacy, or perhaps the same bold or casual attitudes that allows admission of the use of an illicit substance also creates increased self-efficacy, which would be an interesting area for further research.

Demographics

None of the demographic variables influenced self-efficacy. Contrary to conventional wisdom suggesting that older workers with more experience would perceive themselves to be more competent, our study suggests that age is independent of self-efficacy.

Even though women experience large gender inequality in Kenya as per the Global Inequality Index (Bandura and Cervone 1983; Bryant et al. 2011) and although most patients affected with SUDs in Kenya are male (Manghi et al. 2009; National Campaign Against Drug Abuse Authority 2012), gender also did not impact self-efficacy. This is encouraging since most of the primary care in LMIC is delivered by female nurses (International Labour Office 2016; Zurn et al. 2002).

Perception of the Prevalence of Substance Use

Bivariate analysis showed that other variables related to the perceived prevalence of substance use were significantly correlated with self-efficacy. However, only AUD-related perceptions remained significant, possibly due to collinearity or lower rates of observation for OSUD, leading to more frequent interventions and relatively higher AUD-focused confidence. Communicating to health workers the prevalence, health impact, and effective interventions related to the various SUDs might increase care providers’ ability to tackle SUDs other than AUD, as shown by other studies seeking to influence health workers’ attitudes in relation to other conditions (Ballew et al. 2013; Koller 2011; Patten et al. 2012; Pinfold et al. 2003).

Training Needs

The relatively high self-efficacy in screening and intervening for SUDs seems at odds with a prior study showing low level of health workers’ recognition and treatment for SUDs and their complications (Ndetei et al. 2009). The respondents’ self-reported confidence in effective communication and advice around SUDs also conflicts with previous patient-centered studies that suggest a need for PCWs to improve in these areas (Othieno et al. 2013). One explanation could be that the practitioners’ subjective competence is divorced from objective assessments, which is at least anecdotally supported by our research team’s field observations in Kenyan clinics and other studies (Embleton et al. 2013; Jenkins et al. 2010a, b; Muga and Jenkins 2008; Ndetei et al. 2007, 2008a, b, 2009a, b, 2010; Othieno et al. 2000; Nturibi et al. 2009). Based on the classification of stages of learning (from “unconsciously incompetent” to “unconsciously competent”) (Crandall et al. 2003), on the evidence of health workers’ inability to identify and intervene on SUDs (Embleton et al. 2013; Jenkins et al. 2010a, b; Muga and Jenkins 2008; Ndetei et al. 2007, 2008a, b, 2009a, b, 2010; Othieno et al. 2000), and in the context of a self-reported high level of self-efficacy found in our study, our respondents appear to be unconsciously incompetent. This would explain in part why community health workers without prior training in SUD express the same level of confidence in addressing SUD than trained clinicians in our study.

Interestingly, attendance at a TUD- or OSUD-related refresher course did not seem to influence PCW self-efficacy. The refresher courses might be weak in any number of aspects: they may not draw on evidence-based adult learning principles (Davis et al. 1999), may lack support or a sense of mandate to implement the new knowledge (Jenkins et al. 2013; Ndetei et al. 2009), or there may be insufficient complementary changes in skills and attitudes (Hodges et al. 2001). Self-efficacy may be enhanced by providing education on the role of primary care in SUDs, using comprehensive, evidence-based, adult learning techniques, and competency-based learning (Davis et al. 1999; Green and Ellis 1997; Johnstone and Soares 2014) that address knowledge as well as skills and attitudes (Hodges et al. 2001). Our study also indicate that self-efficacy may also be enhanced by having health system decision-makers implement clear support mechanisms for health workers to address SUD. In other words, addressing the social, emotional, and organizational contexts of learning and implementation (Grol and Grimshaw 2003) might be necessary, especially considering the complexity of the Kenyan primary health care system (Kiima and Jenkins 2010). Enlisting support for evidence-based capacity building programs such as eDATA K can address all these aspects simultaneously.

Limitations

Our recruitment process through self-enrollment is prone to selection bias. However, the bias was likely limited, as we had more than 80% of the eligible staff in small facilities (dispensaries, community health centers, and private clinics) and about 50–75% of the eligible staff in larger facilities (counties and sub-counties hospital outpatients departments) participating in this study. The data may have reporting bias due to the stigma associated with SUD, although the proportion of health workers reporting substance use is similar to the Kenyan population (Mokaya et al. 2016). Because the study was conducted in three Kenyan health care regions, in facilities which showed an interest in addressing SUD better, generalization to other regions or countries’ health care workers should be applied with caution. This study only assessed factors that affect the baseline self-efficacy. Further studies should assess whether improved self-efficacy also improves patient outcomes, and which factors augment the improvement of self-efficacy through training.

CONCLUSION

This pioneering study reveals important factors driving self-efficacy for SUDs in LMICs, including the availability of SUD-related support services, perception of SUDs as prevalent in one’s patient population, awareness of the effectiveness of managing SUDs in outpatient settings, risk level of cannabis use, and perceiving one’s professional education having included sufficient training on SUDs. Therefore, providing evidence-based forms of training, addressing the various perceptions of PCW in relation to their roles, real prevalence, and effectiveness of interventions to managing SUDs in outpatient settings, and engaging in system change to provide more support to primary care workers might help increase front-line workers’ self-efficacy and the availability of SUDs-related screenings and interventions.

Acknowledgments

We would like to thank Verena Rossa-Roccor, MD, and Mary Arakelyan for their editorial assistance. Compliance with Ethical Standards

Funding

This study was supported by the Global Mental Health Grant # 0092-04 from Grand Challenges Canada, the Canada Research Chairs program, the Annenberg Physician Training Program in Addiction Medicine (APTPAM), the IMPART clinician fellowship, the Canadian Addiction Medicine Research Fellowship, and the Michael Smith Foundation for Health Research post-doctoral award program.

Footnotes

1

While less well known in certain parts of the world, khat is commonly used in eastern Africa and the Middle East for its stimulant effect (Alem, Kebede, and Kullgren, 1999; Manghi et al. 2009), with lifetime use reported up to 55.7% in Ethiopia (Alem et al. 1999; Cox and Rampes 2003) and 61% in Yemen (Alem et al. 1999; Manghi et al. 2009). In addition to its addictive potential, khat can cause mood dysregulation, psychotic symptoms, and negative impacts on the cardiovascular, respiratory, hepatobiliary, genitourinary, and metabolic systems (Cox and Rampes 2003; National Campaign Against Drug Abuse Authority 2012).

Conflict of Interest

The authors declare that they have no conflict of interest.

Informed Consent

All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1975, as revised in 2000 (5). Informed consent was obtained from all patients for being included in the study.

References

  1. Alem A, Kebede D, Kullgren G. The prevalence and socio-demographic correlates of khat chewing in Butajira, Ethiopia. Acta Psychiatrica Scandinavica. 1999;100(S397):84–91. doi: 10.1111/j.1600-0447.1999.tb10699.x. [DOI] [PubMed] [Google Scholar]
  2. Bahora M, Hanafi S, Chien VH, Compton MT. Preliminary evidence of effects of crisis intervention team training on self-efficacy and social distance. Administration and Policy in Mental Health. 2008;35:159–167. doi: 10.1007/s10488-007-0153-8. [DOI] [PubMed] [Google Scholar]
  3. Ballew P, Castro S, Claus J, Kittur N, Brennan L, Brownson RC. Developing web-based training for public health practitioners: what can we learn from a review of five disciplines? Health Education Research. 2013;28:276–287. doi: 10.1093/her/cys098. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Bandura A, Cervone D. Self-evaluation and self-efficacy mechanisms governing the motivational effects of goal systems. Journal of Personality and Social Psychology. 1983;45:1017–1028. [Google Scholar]
  5. Battersby M, Von Korff M, Schaefer J, Davis C, Ludman E, Greene SM, et al. Twelve evidence- based principles for implementing self-management support in primary care. Joint Commission Journal on Quality & Patient Safety. 2010;36(12):561–570. doi: 10.1016/s1553-7250(10)36084-3. [DOI] [PubMed] [Google Scholar]
  6. Bazzano AT, Zeldin AS, Diab IRS, Garro NM, Allevato NA, Lehrer D, Team, W.R.C.P.O. The healthy lifestyle change program: a pilot of a community-based health promotion intervention for adults with developmental disabilities. American Journal of Preventive Medicine. 2009;37(6 Suppl 1):S201–S208. doi: 10.1016/j.amepre.2009.08.005. [DOI] [PubMed] [Google Scholar]
  7. Beckman H, Hawley S, Bishop T. Application of theory-based health behavior change techniques to the prevention of obesity in children. Journal of Pediatric Nursing. 2006;21(4):266–275. doi: 10.1016/j.pedn.2006.02.012. [DOI] [PubMed] [Google Scholar]
  8. Bryant T, Raphael D, Schrecker T, Labonte R. Canada: a land of missed opportunity for addressing the social determinants of health. Health Policy. 2011;101(1):44–58. doi: 10.1016/j.healthpol.2010.08.022. [DOI] [PubMed] [Google Scholar]
  9. Buckelew SM, Adams SH, Irwin CE, Jr, Gee S, Ozer EM. Increasing clinician self-efficacy for screening and counseling adolescents for risky health behaviors: results of an intervention. Journal of Adolescent Health. 2008;43(2):198–200. doi: 10.1016/j.jadohealth.2008.01.018. [DOI] [PubMed] [Google Scholar]
  10. Cox G, Rampes H. Adverse effects of khat: a review. Advances in Psychiatric Treatment. 2003;9:456–463. [Google Scholar]
  11. Crandall SJ, George G, Marion G, Davis S. Applying theory to the design of cultural competency training for medical students: a case study. Academic Medicine. 2003;78(6):588–594. doi: 10.1097/00001888-200306000-00007. [DOI] [PubMed] [Google Scholar]
  12. Davis D, O’Brien MAT, Freemantle N, Wolf FM, Mazmanian P, Taylor-Vasley A. Impact of formal continuing medical education: do conferences, workshops, rounds, and other traditional continuing education activities change physician behavior or health care out-comes? JAMA. 1999;282(9):867–874. doi: 10.1001/jama.282.9.867. [DOI] [PubMed] [Google Scholar]
  13. Delgadillo J, Moreea O, Outhwaite-Luke H, Dace T, Nicholls B, Ramseyer G, Dale V. Confidence in the face of risk: the risk assessment and management self-efficacy study (RAMSES) Psychiatric Bulletin. 2014;38(2):58–65. doi: 10.1192/pb.bp.112.040394. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. DiClemente CC, Fairhurst SK, Piotrowski NA. Self-efficacy and addictive behaviors. In: Maddux JE, editor. Self-efficacy, adaptation, and adjustment: theory, research, and application. Boston: Springer US; 1995. pp. 109–141. [Google Scholar]
  15. Elder JP, Ayala GX, Harris S. Theories and intervention approaches to health-behavior change in primary care. American Journal of Preventive Medicine. 1999;17(4):275–284. doi: 10.1016/s0749-3797(99)00094-x. [DOI] [PubMed] [Google Scholar]
  16. Embleton L, Atwoli L, Ayuku D, Braitstein P. The journey of addiction: barriers to and facilitators of drug use cessation among street children and youths in western Kenya. PloS One. 2013;8:e53435. doi: 10.1371/journal.pone.0053435. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Franco LM, Bennett S, Kanfer R. Health sector reform and public sector health worker motivation: a conceptual framework. Social Science & Medicine. 2002;54:1255–1266. doi: 10.1016/s0277-9536(01)00094-6. [DOI] [PubMed] [Google Scholar]
  18. Frank E, McLendon L, Elon LK, Denniston M, Fitzmaurice D, Hertzberg Medical students’ self-reported typical counseling practices are similar to those assessed using standardized patients. Medscape General Medicine. 2005;7(1):2. [PMC free article] [PubMed] [Google Scholar]
  19. Gottlieb NH, Mullen PD, McAlister AL. Patients’ substance abuse and the primary care physician: patterns of practice. Addictive Behaviors. 1987;12(1):23–32. doi: 10.1016/0306-4603(87)90005-0. [DOI] [PubMed] [Google Scholar]
  20. Green ML, Ellis PJ. Impact of an evidence-based medicine curriculum based on adult learning theory. J Gen Intern Medicine. 1997;12:742. doi: 10.1046/j.1525-1497.1997.07159.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Grol R, Grimshaw J. From best evidence to best practice: effective implementation of change in patients’ care. Lancet. 2003;362(9391):1225–1230. doi: 10.1016/S0140-6736(03)14546-1. [DOI] [PubMed] [Google Scholar]
  22. Grol R, Wensing M. Barriers to and incentives for achieving evidence-based practice. The Medical Journal of Australia. 2016;180(6 suppl):s57. doi: 10.5694/j.1326-5377.2004.tb05948.x. [DOI] [PubMed] [Google Scholar]
  23. Harris BR, Shaw BA, Sherman BR, Lawson HA. Screening, brief intervention, and referral to treatment for adolescents: attitudes, perceptions, and practice of New York school-based health center providers. Substance Abuse. 2016;37(1):161–167. doi: 10.1080/08897077.2015.1015703. [DOI] [PubMed] [Google Scholar]
  24. Henry-Edwards S, Humeniuk R, Ali R, Monteiro M, Poznyak V. Brief intervention for substance use: a manual for use in primary care. 2003 Retrieved from Geneva. [Google Scholar]
  25. Hodges B, Inch C, Silver I. Improving the psychiatric knowledge, skills, and attitudes of primary care physicians, 1950–2000: a review. The American Journal of Psychiatry. 2001;158(10):1579–1586. doi: 10.1176/appi.ajp.158.10.1579. [DOI] [PubMed] [Google Scholar]
  26. Howe LD, Galobardes B, Matijasevich A, Gordon D, Johnston D, Onwujekwe O, et al. Measuring socio-economic position for epidemiological studies in low- and middle-income countries: a methods of measurement in epidemiology paper. International Journal of Epidemiology. 2012;41(3):871–886. doi: 10.1093/ije/dys037. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Humeniuk R, Ali R, Babor TF, Farrell M, Formigoni ML, Jittiwutikarn J, et al. Validation of the Alcohol, Smoking and Substance Involvement Screening Test (ASSIST) Addiction. 2008;103(6):1039–1047. doi: 10.1111/j.1360-0443.2007.02114.x. [DOI] [PubMed] [Google Scholar]
  28. Humeniuk R, Henry-Edwards S, Ali R, Poznyak V, Monteiro M. The Alcohol, Smoking and Substance Involvement Screening Test (ASSIST): manual for use in primary care. Switzerland: 2010. Retrieved from Geneva. [Google Scholar]
  29. International Labour Office. Women at work: trends 2016. International Labour Office; 2016. [Google Scholar]
  30. Jenkins R, Kiima D, Njenga F, Okonji M, Kingora J, Kathuku D, Lock S. Integration of mental health into primary care in Kenya. World Psychiatry. 2010a;9(2):118–120. doi: 10.1002/j.2051-5545.2010.tb00289.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Jenkins R, Kiima D, Okonji M, Njenga F, Kingora J, Lock S. Integration of mental health into primary care and community health working in Kenya: Context, rationale, coverage and sustainability. Mental Health in Family Medicine. 2010b;7:37–47. [PMC free article] [PubMed] [Google Scholar]
  32. Jenkins R, Othieno C, Okeyo S, Aruwa J, Kingora J, Jenkins B. Health system challenges to integration of mental health delivery in primary care in Kenya—perspectives of primary care health workers. BMC Health Services Research. 2013;13:368. doi: 10.1186/1472-6963-13-368. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Johnstone SM, Soares L. Principles for developing competency-based education programs. Change: The Magazine of Higher Learning. 2014;46(2):12–19. doi: 10.1080/00091383.2014.896705. [DOI] [Google Scholar]
  34. Kadden RM, Litt MD. The role of self-efficacy in the treatment of substance use disorders. Addictive Behaviors. 2011;36(12):1120–1126. doi: 10.1016/j.addbeh.2011.07.032. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Kay K, Shipman C. The confidence code: the science and art of self-assurance—what women should know (1ed.) HarperBusiness 2014 [Google Scholar]
  36. Kiima D, Jenkins R. Mental health policy in Kenya—an integrated approach to scaling up equitable care for poor populations. International Journal of Mental Health Systems. 2010;4:19. doi: 10.1186/1752-4458-4-19. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Koller M. Opening minds at university: results of a contact-based anti-stigma intervention. Mental Health Commission of Canada 2011 [Google Scholar]
  38. Lemieux V, Levesque JF, Ehrmann-Feldman D. Are primary healthcare organizational attributes associated with patient self-efficacy for managing chronic disease? Healthc Policy. 2011;6(4):e89–e105. [PMC free article] [PubMed] [Google Scholar]
  39. Lim SS, Vos T, Flaxman AD, Danaei G, Shibuya K, Adair-Rohani H, AlMazroa MA. A comparative risk assessment of burden of disease and injury attributable to 67 risk factors and risk factor clusters in 21 regions, 1990–2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet. 2012;380:2224–2260. doi: 10.1016/S0140-6736(12)61766-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Manghi RA, Broers B, Khan R, Benguettat D, Khazaal Y, Zullino DF. Khat use: lifestyle or addiction? Journal of Psychoactive Drugs. 2009;41(1):1–10. doi: 10.1080/02791072.2009.10400669. [DOI] [PubMed] [Google Scholar]
  41. Miller Perrin E, Flower KB, Garrett J, Ammerman AS. Preventing and treating obesity: pediatricians’ self-efficacy, barriers, resources, and advocacy. Ambulatory Pediatrics. 2005;5(3):150–156. doi: 10.1367/A04-104R.1. [DOI] [PubMed] [Google Scholar]
  42. Mokaya AG, Mutiso V, Musau A, Tele A, Kombe Y, Ng’ang’a Z, et al. Substance use among a sample of healthcare workers in Kenya: a cross-sectional study. Journal of Psychoactive Drugs. 2016;48(4):310–319. doi: 10.1080/02791072.2016.1211352. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Morgan JF, Killoughery M. Hospital doctors’ management of psychological problems— Mayou & Smith revisited. The British Journal of Psychiatry. 2003;182(2):153–157. doi: 10.1192/bjp.182.2.153. [DOI] [PubMed] [Google Scholar]
  44. Muga FA, Jenkins R. Training, attitudes and practice of district health workers in Kenya. Social Psychiatry and Psychiatric Epidemiology. 2008;43(6):477–482. doi: 10.1007/s00127-008-0327-z. [DOI] [PubMed] [Google Scholar]
  45. National Campaign Against Drug Abuse Authority. Rapid situation assessment of the status of drug and substance abuse in Kenya, 2012. 2012 Retrieved from. [Google Scholar]
  46. Ndetei D, Ongetcha F, Mutiso V, Kuria M, Khasakhala L, Kokonya D. The challenges of human resources in mental health in Kenya. African Journal of Psychiatry. 2007;10(1):33–36. [Google Scholar]
  47. Ndetei D, Khasakhala L, Omolo J. Incentives for health worker retention in Kenya: an assessment of current practice. 2008a Retrieved from Kenya. [Google Scholar]
  48. Ndetei D, Pizzo M, Kuria M, Khasakhala L, Maru M, Mutiso V. Substance abuse and psychiatric co-morbidities: a case study of patients at Mathari Psychiatric Hospital, Nairobi, Kenya. African Journal of Drug and Alcohol Studies. 2008b;7:53–58. doi: 10.4314/ajdas.v7i1.46359. [DOI] [Google Scholar]
  49. Ndetei D, Khasakhala L, Ongecha-Owuor F, Kuria M, Mutiso V, Kokonya D. Prevalence of substance abuse among patients in general medical facilities in Kenya. Substance Abuse. 2009;30(2):182–190. doi: 10.1080/08897070902802125. [DOI] [PubMed] [Google Scholar]
  50. Ndetei D, Mathai M, Khasakhala L, Mutiso V, Mbway A. University medical education in Kenya: the challenge. Medical Teacher. 2010;32:5. doi: 10.3109/0142159X.2010.496011. [DOI] [PubMed] [Google Scholar]
  51. Ndetei DM, Khasakhala LI, Mutiso V, Mbwayo AW. Knowledge, attitude and practice (KAP) of mental illness among staff in general medical facilities in Kenya: practice and policy implications. African Journal of Psychiatry. 2011;14:225–235. doi: 10.4314/ajpsy.v14i3.6. [DOI] [PubMed] [Google Scholar]
  52. Njenga A. Alcohol control and devolution in Kenya 2015 [Google Scholar]
  53. Nturibi EM, Akinsola AK, McCurdy SA. Smoking prevalence and tobacco control measures in Kenya, Uganda, the Gambia and Liberia: a review. The international journal of tuberculosis and lung disease: the official journal of the International Union against Tuberculosis and Lung Disease. 2009;13(2):165–170. [PubMed] [Google Scholar]
  54. O’Campo P, Kirst M, Tsamis C, Chambers C, Ahmad F. Implementing successful intimate partner violence screening programs in health care settings: evidence generated from a realist-informed systematic review. Social Science & Medicine. 2011;72(6):855–866. doi: 10.1016/j.socscimed.2010.12.019. [DOI] [PubMed] [Google Scholar]
  55. Othieno CJ, Kathuku DM, Ndetei DM. Substance abuse in outpatients attending rural and urban health centres in Kenya. East African Medical Journal. 2000;77(11):592–595. doi: 10.4314/eamj.v77i11.46728. [DOI] [PubMed] [Google Scholar]
  56. Othieno C, Jenkins R, Okeyo S, Aruwa J, Wallcraft J, Jenkins B. Perspectives and concerns of clients at primary health care facilities involved in evaluation of a national mental health training programme for primary care in Kenya. International Journal of Mental Health Systems. 2013;7:5. doi: 10.1186/1752-4458-7-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Ozer EM, Adams SH, Gardner LR, Mailloux DE, Wibbelsman CJ, Irwin CE., Jr Provider self-efficacy and the screening of adolescents for risky health behaviors. Journal of Adolescent Health. 2004;35(2):101–107. doi: 10.1016/j.jadohealth.2003.09.016. [DOI] [PubMed] [Google Scholar]
  58. Patten SB, Remillard A, Phillips L, Modgill G, Szeto AC, Kassam A, Gardner DM. Effectiveness of contact-based education for reducing mental illness-related stigma in pharmacy students. BMC Medical Education. 2012;12:120. doi: 10.1186/1472-6920-12-120. [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Pinfold V, Toulmin H, Thornicroft G, Huxley P, Farmer P, Graham T. Reducing psychiatric stigma. The British Journal of Psychiatry. 2003;182:342–346. doi: 10.1192/bjp.182.4.342. [DOI] [PubMed] [Google Scholar]
  60. Republic of Kenya. Health sector strategic and investment plan (KHSSP) July 2013–June 2017: the second medium term plan for health. Ministry of Medical Services and Ministry of Public Health & Sanitation 2013 [Google Scholar]
  61. Saxena S, Thornicroft G, Knapp M, Whiteford H. Resources for mental health: scarcity, inequity, and inefficiency. The Lancet. 2007;370:878–889. doi: 10.1016/S0140-6736(07)61239-2. [DOI] [PubMed] [Google Scholar]
  62. Schram P, Harris SK, Van Hook S, Forman S, Mezzacappa E, Pavlyuk R, Levy S. Implementing adolescent Screening, Brief Intervention, and Referral to Treatment (SBIRT) education in a pediatric residency curriculum. Substance Abuse. 2015;36(3):332–338. doi: 10.1080/08897077.2014.936576. [DOI] [PubMed] [Google Scholar]
  63. Schrecker T, Labonte R. Taming the brain drain: a challenge for public health systems in southern Africa. Internal Journal of Occupational and Environmental Health. 2004;10(4):409–415. doi: 10.1179/oeh.2004.10.4.409. [DOI] [PubMed] [Google Scholar]
  64. Staff of the Global Legal Research, D. Legal status of Khat in selected jurisdictions. 2015 Retrieved from https://www.loc.gov/law/help/legal-status-of-khat/legal-status-khat.pdf.
  65. The Academy of Medical, S. Challenges and priorities for global mental health research in low- and middle-income countries: symposium report. 2008 Retrieved from. [Google Scholar]
  66. Thompson SC, Schwankovsky L, Pitts J. Counselling patients to make lifestyle changes: the role of physician self-efficacy, training and beliefs about causes. Family Practice. 1993;10(1):70–75. doi: 10.1093/fampra/10.1.70. [DOI] [PubMed] [Google Scholar]
  67. Tumwine J. Implementation of the framework convention on tobacco control in Africa: current status of legislation. International Journal of Environmental Research and Public Health. 2011;8(11):4312–4331. doi: 10.3390/ijerph8114312. [DOI] [PMC free article] [PubMed] [Google Scholar]
  68. WHO. mhGAP: Mental Health Gap Action Programme: scaling up care for mental, neurological, and substance use disorders. World Health Organization; 2008. [PubMed] [Google Scholar]
  69. WHO. ATLAS on substance use 2010—resources for the prevention and treatment of substance use disorders. World Health Organization; 2010. [Google Scholar]
  70. WHO. Human resources for mental health: workforce shortages in low- and middle-income countries. 2011 Retrieved from. [Google Scholar]
  71. Woddruff SI, Candelaria JI, Elder JP. Recruitment, training outcomes, retention, and performance of community health advisors in two tobacco control interventions for Latinos. Journal of Community Health. 2010;35(2):124–134. doi: 10.1007/s10900-009-9207-z. [DOI] [PubMed] [Google Scholar]
  72. Zurn P, Dal Poz M, Stilwell B, Adams O. Imbalances in the workforce: briefing paper. World Health Organization; 2002. [Google Scholar]

RESOURCES