Determinants of the Intention of Senegal's Physicians to Use Telemedicine in Their Professional Activities

Abstract

Background:

In Senegal, physicians are unevenly distributed, leading to unequal access to healthcare in underserved areas. Telemedicine is seen as a potential means to address this problem.

Introduction:

Telemedicine's potential to improve access depends, in part, on physicians' intention to use the technology. In Senegal that intention is not well known. This study aimed to determine that intention and the factors that influence it.

Materials and Methods:

We conducted a cross-sectional study between January and February 2015 with a random sample of 168 physicians working in public hospitals and 153 in district health centers in Senegal. Data were collected using two questionnaires and analyzed using descriptive statistics, correlations, and linear regression.

Results:

The intention to use telemedicine by physicians working in public hospitals and district health centers was moderate and was positively correlated with their attitude, subjective norm, and perceived behavioral control. The intention of the physicians working in public hospitals was also positively correlated with their region and status as contract employee, but negatively with their status as government employee. That of the physicians working in district health centers was negatively correlated with their age and years of practice.

Discussion:

These results showed that, overall, the intention of Senegal's physicians to use telemedicine was moderate and could be improved by acting on factors related to their perceived behavioral control and other factors correlated with their intention.

Conclusions:

Physicians' intention to use telemedicine in Senegal is fair but could be improved by addressing the factors identified in this study.

Introduction

In Senegal, more than 60% of physicians work in Dakar, the capital city, but it comprises only 23% of the national population.¹ This uneven distribution leads to inequitable access to healthcare.² One of the main causes of this misdistribution is difficulty in recruiting and retaining physicians outside Dakar. This difficulty can be attributed to many factors; those well-known being occupational, personal, family, and community factors.³ International migration may also exacerbate these distribution issues with data showing that more than 51% of Senegalese-trained physicians work abroad.⁴ Senegal has adopted a number of measures to promote the recruitment and retention of physicians outside Dakar, including the increase of the production of physicians through the creation of new medical schools and the improvement of the management of these physicians.³ To date, however, these measures have not resulted in improved equitable distribution of physicians between Dakar and the rest of the country.

Telemedicine, which can be defined as the practice of medicine from a distance through the use of information and communication technologies, is seen as a potentially effective way of increasing access to healthcare in underserved areas. It would do so by improving the recruitment and retention of health professionals in these areas with technology-mediated linkages to specialist services in urban centers.^5

–8 During the past decade, several telemedicine projects were implemented in Senegal⁹ although many of them have not progressed past an early stage.¹⁰ These failures challenge the expectations of the benefits of telemedicine.

The determinants of these failures are not well known. According to Broens et al.,¹¹ the factors that determine the success or failure of telemedicine projects can be categorized as individual, technological, financial, organizational, political, and legislative. This study focused specifically on the individual factors. It examined the intention of Senegal's physicians to use telemedicine in their professional activities and identified the factors influencing that intention. It is based on an adaptation of the theory of planned behavior (TPB), which is known to be effective in studying intentions of health professionals.^12,13 It describes how intention, which provides the needed motivation for a person to implement a new behavior, is the proximal determinant of behavior.^12,13 Intention depends on one's attitude toward the behavior, subjective norm, and perceived behavioral control. Attitude in this study refers to the degree to which the performance of the behavior is positively or negatively valued.¹² Subjective norm refers to the perceived social pressure to either engage in or avoid the behavior.¹² Perceived behavioral control refers to individuals' perceptions of their ability to perform the behavior.^12,14,15 These three constructs depend, respectively, on behavioral, normative, and control beliefs. Behavioral belief refers to the subjective probability that the behavior will produce a given outcome. Normative beliefs refer to the perceived behavioral expectations of important referents, including individuals or groups. Control beliefs refer to the perceived presence of factors that may facilitate or impede the performance of the behavior.¹⁴ These beliefs are influenced by individuals' characteristics and the context in which the behavior is adopted. This theory has been used previously to study the determinants of telemedicine use and has provided satisfactory results.^16,17

Materials and Methods

Study Design

A cross-sectional study was conducted between January and February 2015 in Senegal. This design is a descriptive approach commonly used in epidemiology.¹⁸ It consists of measuring a specific outcome and factors associated with it at a specific point in time. It is also relatively inexpensive and does not require much time to be conducted or for researchers to follow-up with participants since they are surveyed only once.¹⁹

Population, Sampling, and Recruitment

The study population consisted of physicians working in public hospitals (n = 596) and district health centers (n = 187). This study did not include physicians working in the private sector due to the limited time frame, the difficulty of recruiting physicians in the private sector, and the research costs associated with the inclusion of these physicians. Among the 596 physicians working in public hospitals, 200 were randomly selected for inclusion, while all 187 physicians working in district health centers were invited to participate. The physicians were contacted to establish their availability, willingness to participate, and consent to respond to the questionnaire. The final sample consisted of 168 physicians working in public hospitals and 153 physicians working in district health centers.

Data Collection

Standardized survey questionnaires were administered in person by the lead researcher and three trained investigators. Data collection was carried out in hospitals, district health centers, hotels, conference rooms, and physicians' homes, depending on each physician's availability and preference. In cases of missing data, physicians were called to either complete or correct the information.

Instruments

Data were collected using two questionnaires (see Appendices A1 and A2). The first was developed specifically for the physicians working in public hospitals, while the second was conceived for the physicians working in district health centers. Each of the two questionnaires was composed of two parts. The first part solicited physicians' sociodemographic and professional characteristics, and the second asked questions about physicians' intention, attitude, subjective norm, and perceived behavioral control. The items used to assess physicians' intention were informed by the literature, while those used to assess their attitude, subjective norm, and perceived behavioral control were developed by studying their positive and negative behavioral, normative, and control beliefs. These items were content validated with a group of seven experts from the University of Ottawa and Laval University to revise, add, or remove certain items. The final number of items for each of the four theoretical constructs (intention, attitude, subjective norm, and perceived behavioral control) varied from 3 to 11 for the first questionnaire (public hospitals) and from 3 to 10 for the second questionnaire (district health centers). Tables 1 –4 show the number of items of each theoretical construct in each questionnaire. These items were measured using a 7-point Likert scale. We scored each item by each physician, then by a group of physicians. Next, we scored each construct by a group of physicians. Tables 1 –4 show the mean score and standard deviation of each item and construct by a group of physicians. Later, we calculated Cronbach's alpha coefficients. These coefficients are shown in Table 5 and indicate satisfactory internal consistency for all constructs and both groups of physicians.

Table 1.

Descriptive Analysis of Physicians' Intention

	PHYSICIANS WORKING IN PUBLIC HOSPITALS		PHYSICIANS WORKING IN DISTRICT HEALTH CENTERS
ITEM	MEAN	SD	MEAN	SD
Intention	4.29	1.77	4.28	1.69
Over the 12 coming months, I intend to use telemedicine in my professional activities.	4.99	1.91	5.19	1.73
Over the 12 coming months, I will use telemedicine in my professional activities.	4.39	2.03	4.21	1.99
Over the 12 coming months, the probability that I use telemedicine in my professional activities is very low or very high.	3.50	2.11	3.45	2.03

The mean score was considered very unfavorable when it ranged between 1 and 2, unfavorable when it ranged between 2 and 3, slightly unfavorable when it ranged between 3 and 4, moderate when it ranged between 4 and 5, slightly favorable when it ranged between 5 and 6, favorable when it ranged between 6 and 7, and very favorable when it was 7.

SD, standard deviation.

Table 2.

Descriptive Analysis of Physicians' Attitude

	PHYSICIANS WORKING IN PUBLIC HOSPITALS		PHYSICIANS WORKING IN DISTRICT HEALTH CENTERS
ITEM	MEAN	SD	MEAN	SD
Attitude	5.93	0.81	5.80	0.92
To me, using telemedicine in my professional activities over the 12 coming months is very useless or very useful.	6.26	0.81	6.18	1.22
To me, using telemedicine in my professional activities over the 12 coming months is very harmful or very beneficial.	6.25	0.79	6.23	0.95
To me, using telemedicine in my professional activities over the 12 coming months is very imprudent or very prudent.	5.78	1.12	5.61	1.22
Using telemedicine in my professional activities over the 12 coming months would allow me not to evacuate some of my patients.	5.43	1.62	5.66	1.46
Using telemedicine in my professional activities over the 12 coming months would allow me to gain time.	6.08	1.07	5.92	1.36
Using telemedicine in my professional activities over the 12 coming months would allow me to avoid medical errors.	5.75	1.39	5.93	1.31
Using telemedicine in my professional activities over the 12 coming months would allow me to train myself professionally.	6.10	1.59	5.84	1.67
Using telemedicine in my professional activities over the 12 coming months would allow me to treat my patients at distance.	5.49	1.83	5.14	1.90
Using telemedicine in my professional activities over the 12 coming months would allow me to get the experts' opinions at distance.	6.31	1.44	6.07	1.70
Using telemedicine in my professional activities over the 12 coming months would allow me to better organize the evacuation of my patients.	6.15	1.22	N/A	N/A
Using telemedicine in my professional activities over the 12 coming months would allow me to contribute to the training of other health professionals.	5.61	1.85	N/A	N/A
Using telemedicine in my professional activities over the 12 coming months would allow me to reduce the expenses incurred by my patients.	N/A	N/A	5.44	1.86

N/A, not applicable.

Table 3.

Descriptive Analysis of Physicians' Subjective Norm

	PHYSICIANS WORKING IN PUBLIC HOSPITALS		PHYSICIANS WORKING IN DISTRICT HEALTH CENTERS
ITEM	MEAN	SD	MEAN	SD
Subjective norm	5.95	0.76	5.81	0.82
My Health Ministry would strongly approve or disapprove that I use telemedicine in my professional activities over the 12 coming months.	6.12	0.93	6.08	1.09
My population would strongly approve or disapprove that I use telemedicine in my professional activities over the 12 coming months.	5.85	1.03	5.67	1.26
My colleagues (physicians) would strongly approve or disapprove that I use telemedicine in my professional activities over the 12 coming months.	6.09	0.92	5.71	1.26
My patients would strongly approve or disapprove that I use telemedicine in my professional activities over the 12 coming months.	5.68	1.21	5.56	1.27
My paramedical staff would strongly approve or disapprove that I use telemedicine in my professional activities over the 12 coming months.	5.95	0.97	5.54	1.12
My students would strongly approve or disapprove that I use telemedicine in my professional activities over the 12 coming months.	6.33	0.80	N/A	N/A
The administrators of my hospital would strongly approve or disapprove that I use telemedicine in my professional activities over the 12 coming months.	5.67	1.33	N/A	N/A
The head physician of my medical region would strongly approve or disapprove that I use telemedicine in my professional activities over the 12 coming months.	N/A	N/A	5.94	1.06
The academics would strongly approve or disapprove that I use telemedicine in my professional activities over the 12 coming months.	N/A	N/A	6.05	0.98
The financial and technical partners of my structure would strongly approve or disapprove that I use telemedicine in my professional activities over the 12 coming months.	N/A	N/A	5.93	1.04

N/A, not applicable.

Table 4.

Descriptive Analysis of Physicians' Perceived Behavioral Control

	PHYSICIANS WORKING IN PUBLIC HOSPITALS		PHYSICIANS WORKING IN DISTRICT HEALTH CENTERS
ITEM	MEAN	SD	MEAN	SD
Perceived behavioral control	5.47	0.81	4.99	0.95
To me, using telemedicine in my professional activities over the 12 coming months will be very difficult or very easy.	4.53	1.85	4.12	1.87
If I want, I can easily use telemedicine in my professional activities over the 12 coming months.	4.12	2.10	3.68	2.09
I feel that I will be able to use telemedicine in my professional activities over the 12 coming months.	5.34	1.69	5.60	1.50
I will use telemedicine in my professional activities over the 12 coming months if the security of my patients' data is guaranteed.	6.52	0.84	6.28	1.04
I will use telemedicine in my professional activities over the 12 coming months despite my time constraints.	5.62	1.33	5.32	1.48
I will use telemedicine in my professional activities over the 12 coming months if I'm able to communicate with the respondents.	6.28	0.90	5.75	1.49
I will use telemedicine in my professional activities over the 12 coming months if I'm able to manipulate telemedicine equipment.	3.74	2.06	3.42	1.88
I will use telemedicine in my professional activities over the 12 coming months if I'm trained to use it.	6.52	0.77	N/A	N/A
I will use telemedicine in my professional activities over the 12 coming months if I'm able to protect my patients	6.56	0.65	N/A	N/A
I will use telemedicine in my professional activities over the 12 coming months if I'm able to get the adhesion of my population.	N/A	N/A	5.21	1.56
I will use telemedicine in my professional activities over the 12 coming months if I'm able to get the informed consent of my patients.	N/A	N/A	5.53	1.50

N/A, not applicable.

Table 5.

Cronbach's Alpha Coefficients

CONSTRUCTS OF TPB	PHYSICIANS WORKING IN HOSPITALS	PHYSICIANS WORKING IN DISTRICT HEALTH CENTERS
Intention	0.85	0.85
Attitude	0.81	0.82
Subjective norm	0.85	0.87
Perceived behavioral control	0.72	0.75

TPB, theory of planned behavior.

Data Analysis

We performed analyses using SPSS version 23. We conducted descriptive analyses of physicians' characteristics, intention, attitude, subjective norm, and perceived behavioral control. During these analyses, we calculated the frequency of each characteristic of physicians and estimated the mean scores by sex, age, region of practice, employment status, specialization, number of years in practice, and number of years in the current position.

Next, we conducted correlation analyses to explore the relationship between intention and the three other theoretical constructs (attitude, subjective norm, and perceived behavioral control) on the one hand and between intention and sociodemographic and professional characteristics on the other. In these analyses, we used Pearson's correlation coefficients to investigate the relationships between intention and the three other theoretical constructs, and Spearman's correlation coefficients to look at the relationships between intention and physicians' sociodemographic and professional characteristics. Pearson correlation coefficient measures the strength of the linear relationship between normally distributed variables. When variables are not normally distributed or the relationship between the variables is not linear, it is more appropriate to use the Spearman correlation coefficient.²⁰

Finally, we performed a two-step multiple regression analysis. In the first step, we introduced only the two theoretical constructs (attitude, subjective norm, and perceived behavioral control) into the model. Afterward, we added the characteristics of physicians that were significantly correlated with intention at p < 0.05, namely region, government employee status, and contract employee status for the physicians working in hospitals and age and number of years in practice for the physicians working in district health centers. For each variable, the criteria for inclusion and retention in the final models were fixed at p < 0.10 (inclusion) and p < 0.05 (retention). Before performing the regression analysis, we checked for outliers and respected the postulates of linear regression (namely, the linearity of the relationship between intention and the various constructs of the TPB and the normality of errors). For physicians working in public hospitals, we used Mahalanobis distance to identify three outliers that we took out of the database. The number of physicians working in public hospitals who were included in the final analysis was 165. We did not find any multicollinearity, all kurtosis statistics were below 2, but attitude skewness was 1. Square root transformation did not eliminate the skewness score, so we retained its original value in the regression model for ease of interpretation. For physicians working in district health centers, we took two outliers out of the database using Mahalanobis distance. The number of physicians working in district health centers who were included in the final analysis was 151. We did not find any multicollinearity; all skewness statistics were below 1 and all kurtosis statistics below 2.

Results

Descriptive Analysis of Physicians' Characteristics

In total, 165 physicians working in public hospitals and 151 physicians working in district health centers were included in the analyses. Table 6 shows the distribution of these physicians by sex, age, region of practice, specialization, employment status, number of years in medical practice, and number of years in the current post. The 165 physicians working in public hospitals were selected from 18 public hospitals spread across eight administrative regions, including Dakar, Diourbel, Kaffrine, Louga, Saint-Louis, Thiès, Fatick, and Kaolack. The 151 physicians working in district health centers were selected in 70 district health centers spread across the 14 administrative regions of Senegal. In the two groups, the physicians working in Dakar were more numerous.

Table 6.

Sociodemographic Characteristics of Physicians

	PHYSICIANS WORKING IN HOSPITALS		PHYSICIANS WORKING IN DISTRICT HEALTH CENTERS
CHARACTERISTICS	n	%	n	%
Sex
Male	112	67.88	97	64.24
Female	53	32.12	54	35.76
Age
≤30	10	6.10	16	10.59
31–35	39	23.64	38	25.16
36–40	46	27.88	45	29.80
41–45	26	15.76	20	13.24
46–50	22	13.33	18	11.92
>50	22	13.33	14	9.27
Medical regions
Dakar	125	75.76	100	66.23
Out of Dakar	40	24.24	51	33.77
Specialization
General practitioners	30	18.18	64	42.38
Specialist physicians	135	81.82	74	49.01
Trainees physicians	0	0.00	13	8.61
Employment status
Government employee	112	67.88	96	63.58
Contract employee	38	23.03	42	27.81
Trainee physicians	15	9.09	13	8.61
No. of years in medical practice
≤5	32	19.39	48	31.79
6–10	50	30.30	47	31.13
11–15	44	26.67	36	23.84
16–20	19	11.51	13	8.61
>20	20	12.12	7	4.63
No. of years in their current center or hospital
≤5	68	41.21	126	83.44
6–10	56	33.94	23	15.23
11–15	26	15.76	2	1.32
16–20	7	4.24	0	0.00
>20	8	4.85	0	0.00

The physicians working in district health centers were younger (39 years) than their colleagues who worked in public hospitals (41 years). Female physicians represented 32.5% of the physicians working in public hospitals and 35.76% of the physicians working in district health centers. In the two groups, specialist physicians and government employees were also more numerous than general practitioners, trainee physicians, and contract employees. The number of years of practice was higher among the physicians working in public hospitals (12 ± 7) than among the physicians working in district health centers (9 ± 6). Similarly, the number of years in the current post was higher among the physicians working in public hospitals (8 ± 6) than among the physicians working in district health centers (3 ± 3).

Descriptive Analysis of the Constructs of the TPB

The average intention of both the physicians working in public hospitals and district health centers to use telemedicine in their professional activities was moderate. Their average attitude and subjective norm were slightly favorable. The average perceived behavioral control of the physicians working in public hospitals was slightly favorable, while that of the physicians working in district health centers was moderate (Tables 1 –4).

Correlations Between Intention and Theoretical Constructs

Physicians working in public hospitals

We calculated Pearson's correlation between intention and the three theoretical constructs (attitude, subjective norm, and perceived behavioral control) corresponding to this group of physicians and between the three theoretical constructs themselves. All correlations were significant at p < 0.01, indicating a significant statistical relationship between intention and the three theoretical constructs and between the three theoretical constructs themselves (Table 7).

Table 7.

Correlations Between Intention and Theoretical Constructs

GROUP OF PHYSICIANS	SCALE	INTENTION	ATTITUDE	SUBJECTIVE NORM	PERCEIVED BEHAVIORAL CONTROL
Physicians working in public hospitals	Intention	1
	Attitude	0.348^*	1
	Subjective norm	0.262^*	0.521^*	1
	Perceived behavioral control	0.607^*	0.404^*	0.453^*	1
Physicians working in district health centers	Intention	1
	Attitude	0.349^*	1
	Subjective norm	0.247^*	0.484^*	1
	Perceived behavioral control	0.663^*	0.454^*	0.432^*	1

Correlation is significant at 0.01 (bilateral).

Physicians working in district health centers

We performed the same tests for this group of physicians. Again, all correlations were significant at p < 0.01 (Table 7).

Correlation Between Intention and Physicians' Characteristics

Physicians working in public hospitals

We calculated Spearman's correlation between intention and physicians' characteristics namely their sex, age, region of practice, specialization, employment status, number of years in medical practice, and number of years in their current post (Table 8). Only the correlations between intention and region of practice, government employee status, and contract employee status were significant (p < 0.05). Consequently, the regression model for this group of physicians included only these three characteristics.

Table 8.

Correlation Between Intention and Physicians' Characteristics

GROUP OF PHYSICIANS		INTENTION	REGION	AGE	SEX	SPECIALIST	GENERALIST	YEARS IN CURRENT HOSPITAL	YEARS IN PRACTICE	GOVERNMENT EMPLOYEE	CONTRACT EMPLOYEE	TRAINEE DOCTORS
Physicians working in public hospitals	Intention	1
	Region	0.192^*	1
	Age	−0.103	−0.087	1
	Sex	−0.143	−0.056	0.008	1
	Specialists	−0.117	−0.283^**	0.444^**	−0.080	1
	Generalists	0.117	0.283^**	−0.444^**	0.080	−1.000^**	1
	Years in current hospital	−0.008	−0.359^**	0.639^**	−0.040	0.527^**	−0.527^**	1
	Years in practice	−0.128	−0.295^**	0.727^**	−0.092	0.491^**	−0.491^**	0.791^**	1
	Government employee	−0.208^**	−0.277^**	0.232^**	−0.083	0.450^**	−0.450^**	0.389^**	0.318^**	1
	Contract employee	0.171^*	0.318^**	−0.006	0.014	−0.093	0.093	−0.172^*	−0.157^*	−0.768^**	1
	Trainee physicians	0.071	0.018	−0.368^**	0.099	−0.562^**	0.562^**	−0.383^**	−0.287^**	−0.460^**	−0.167^*	1
Physicians working in district health centers	Intention	1
	Region	0.000	1
	Age	−0.168^*	−0.140	1
	Sex	−0.045	−0.344^**	−0.074	1
	Specialists	−0.084	−0.234^**	0.477^**	0.052	1
	Generalists	0.036	0.168^*	−0.233^**	−0.063	−0.841^**	1
	Years in current center	0.054	−0.068	0.575^**	−0.028	0.353^**	−0.142	1
	Years in practice	−0.178^*	−0.134	0.884^**	−0.072	0.545^**	−0.292^**	0.555^**	1
	Government employee	0.033	0.245^**	0.428^**	−0.124	0.347^**	−0.204^*	0.333^**	0.465^**	1
	Contract employee	0.026	−0.179^*	−0.319^**	−0.007	−0.212^**	0.158	−0.185^*	−0.353^**	−0.753^**	1
	Trainee physicians	−0.021	−0.080	−0.326^**	0.165^*	−0.250^**	0.067	−0.394^**	−0.359^**	−0.405^**	−0.120	1

Correlation is significant at 0.05 (bilateral).

Correlation is significant at 0.01 (bilateral).

Physicians working in district health centers

We performed the same test for this group of physicians. Only the correlations between intention and physicians' age and number of years in practice were significant (p < 0.05) (Table 8). Consequently, the regression model included only age and number of years in medical practice.

Regression Analysis

Physicians working in hospitals

Table 9 shows the results of the regression analysis for this group of physicians. The results indicate that the intention of these physicians to use telemedicine in their professional activities is influenced by perceived behavioral control (p < 0.0001), that is, their perception of their ability to use telemedicine in their professional activities. The final regression models explain 36.80% of the variance in their intention.

Table 9.

Regression Coefficients

		NONSTANDARDIZED COEFFICIENTS		STANDARDIZED COEFFICIENTS
GROUP OF PHYSICIANS	MODEL	A	SE	BETA	T	p
Physicians working in public hospitals	Constant	−2.788	1.043		−2.672	0.008
	Attitude	0.293	0.162	0.134	1.804	0.073
	Subjective norm	−0.189	0.176	−0.081	−1.071	0.286
	Perceived behavioral control	1.248	0.153	0.573	8.155	0.000
	Region	0.373	0.268	0.090	1.390	0.166
	Contract employee	−0.215	0.416	−0.050	−0.517	0.606
	Government employee	−0.586	0.362	−0.155	−1.621	0.107
Physicians working in district health centers	Constant	−1.083	1.056		−1.026	0.307
	Attitude	0.242	0.133	0.133	1.813	0.072
	Subjective norm	−0.089	0.148	−0.044	−0.604	0.547
	Perceived behavioral control	1.098	0.126	0.616	8.680	0.000
	Age	−0.017	0.028	−0.074	−0.603	0.548
	No. of years in practice	−0.037	0.034	−0.132	−1.105	0.271

Dependent variable: intention.

SE, standard error.

Physicians working in district health centers

Table 9 shows the results of regression analysis for this group of physicians. The results show that the intention of physicians in this group to use telemedicine in their professional activities is similarly influenced by perceived behavioral control (p < 0.0001). The final regression models explain 43.90% of the variance in their intention.

Discussion

Our results show that the intention of the physicians working in Senegal's public hospitals and district health centers was moderate and positively correlated with their attitude, subjective norm, and perceived behavioral control. These results are consistent with other findings.^16,17 For example, they corroborate the results of Hu and Chau¹⁷ that indicate a moderate intention for the physicians working in public tertiary hospitals and those of Kuo et al.¹⁶ that show a positive correlation between physicians' intention and their attitude, subjective norm, and perceived behavioral control. These results suggest that the intention of Senegal's physicians regarding the use of telemedicine need to be improved and can be influenced by acting on their attitude, subjective norm, and perceived behavioral control.

The intention of the physicians working in public hospitals was also positively correlated with their medical region and their status as contract employees and negatively correlated with their government employee status. Concerning the physicians working in district health centers, their intention was found to be negatively correlated with their age and number of years in practice. These results suggest that the intention of these physicians is influenced by some of their sociodemographic and professional characteristics. Senegal's health authorities should consider these characteristics when formulating strategies to improve telemedicine adoption in public hospitals and district health centers.

The intention of the physicians working in public hospitals was explained, at 36.80%, by their perceived behavioral control, while the intention of the physicians working in district health centers was explained, at 43.90%, by the same psychosocial determinant. These results are consistent with the findings of other researchers.^16,17,21 They suggest that Senegal's health authorities should act on physicians' perceived behavioral control by acting on their control beliefs, particularly their beliefs regarding their ability to manage their time, communicate with their respondents when they need them, guarantee the security of their patients' data, easily use the technology, get the adhesion of their population, secure their patients' informed consent, and maintain telemedicine equipment. Other researchers have focused on these beliefs and found that they can be a barrier to the use of telemedicine. For example, lack of time was seen as a barrier to the use of telemedicine,^22,23 as well as concerns related to data security,²⁴ ease-of-use of technologies,^25,26 population adhesion,²⁷ and patients' informed consent.^28,29 Acting on these beliefs could improve physicians' perception of their ability to use telemedicine. Consequently, influencing changes in these beliefs could improve both physicians' intention and their behavior regarding the use of telemedicine. Finally, these changes could positively influence physicians' recruitment and retention where there are current shortages and maldistribution, thereby resulting overtime in improved healthcare access for the population.

Conclusions

The findings from this study revealed that the intention of the physicians working in both public hospitals and district health centers was moderate and influenced by their perceived behavioral control. Physician intention was correlated with certain factors, meaning that it could improve with changes in these factors. Acting on these factors to increase physicians' perceived behavioral control could lead successively to improved intention, use of telemedicine, recruitment, and retention outside Dakar, access to healthcare, and population health. These results provide useful information to Senegal's decision-makers, planners, and researchers working on improving telemedicine use and rural physician recruitment and retention.

Footnotes

Acknowledgments

The authors thank the program of population health of the University of Ottawa and the health institutions in Senegal for their technical support. In addition, they also thank Prof. Mbayang Ndiaye Niang, Dr. Myuri Manogaran, Dr. Mahamane Mariko, Sophie Muluka, Marie Claude Lavoie, and Touh Diarra for their social and technical support. This research was financed by the Faculty of Graduate and Postdoctoral Studies (FGPS) of the University of Ottawa, CIHR through its grant (No. 106493) related to the project entitled “Source Country perspectives on the migration of highly trained health personnel: causes, consequences, and responses,” Telfer School of Management Research Fund (SMRF), and Global Health Practicum Grant (GHPG).

Ethics Approval and Consent to Participate

This study required approval from the Research Ethics Board of the University of Ottawa (No. H 09-13-12) and Senegal's National Ethics Committee on Health Research (No. 205 MSAS/DPRS/CNERS). It also required authorization from the Ministry of Health of Senegal (No. 111 MSAS/DPRS/DR). Each participant provided his or her written informed consent, and the anonymity of their information was respected during the whole research process.

Authors' Contributions

The lead author designed the research project, developed the two questionnaires, conducted the validation of these questionnaires, administered the questionnaires, built databases, analyzed data, drafted the first version, and incorporated feedback from coauthors on the first and subsequent drafts of this article. E.K. supervised questionnaires' validation and data analysis. R.L. and I.L.B. reviewed the research project, participated in the validation of the two questionnaires, supervised the writing process, and reviewed the various versions of this article.

Disclosure Statement

No competing financial interests exist.

Appendix A1

Questionnaire used to study the intention of the physicians working in public hospitals to use telemedicine in their professional activities

Ce questionnaire est destiné à l'étude de l'intention des médecins des hôpitaux publics du Sénégal d'utiliser la télémédecine dans leurs activités professionnelles. Nous vous prions de bien vouloir le remplir et nous le retourner, dans les deux semaines qui suivent sa réception, à l'adresse bly022@uottawa.ca ou apholyca@hotmail.com ou à la direction régionale de la santé. La confidentialité et l'anonymat de vos informations seront rigoureusement respectés.

Appendix A2

Questionnaire used to study the intention of the physicians working in district health centers to use telemedicine in their professional activities

Ce questionnaire est destiné à l'étude de l'intention des médecins des centres de santé de district du Sénégal d'utiliser la télémédecine dans leurs activités professionnelles. Nous vous prions de bien vouloir le remplir et nous le retourner, dans les deux semaines qui suivent sa réception, à l'adresse bly022@uottawa.ca ou apholyca@hotmail.com ou à la direction régionale de la santé. La confidentialité et l'anonymat de vos informations seront rigoureusement respectés.

References

Zurn

, Codjia

, Sall

, Braichet

J-M

. How to recruit and retain health workers in underserved areas: The Senegalese experience. Bull World Health Organ, 2010; 88:386–389.

Agence Nationale de la Statistique et de la Démographie (ANSD) et ICF International. Multiple indicators demographic and health survey Senegal (EDS-MICS) 2010–2011 (in French). Rapport final. Calverton, MD: ANSD et ICF International, 2012; 115–273.

Zurn

, Codjia

, Sall

. Retention of health workers in difficult areas in Senegal, increasing access to health workers in rural and remote areas: Case Study 1 (in French). Geneva, Switzerland: World Health Organization, 2010; 10–11.

Clemens

, Pettersson

. New Data on African Health Professionals Abroad: Working Paper 95. Washington, DC: Center for Global Development, 2007; 1–23.

Ellenby

, Marcin

. The role of telemedicine in pediatric critical care. Crit Care Clin, 2015; 31:275–290.

Duplantie

, Gagnon

M-P

, Fortin

J-P

, Landry

. Telehealth and the recruitment and retention of physicians in rural and remote regions: A Delphi study. Can J Rural Med, 2007; 12:30–36.

Fortin

, Landry

, Gagnon

, Duplantie

, Bergeron

, Galarneau

, et al. Telehealth: a strategy to support the practice of physicians in remote areas; research report (in French). Ottawa, ON: Fondation Canadienne de Recherche sur les Services de Santé, 2006; 9–10.

Gagnon

, Duplantie

, Fortin

, Landry

. Exploring the effects of telehealth on medical human resources supply: A qualitative case study in remote regions. BMC Health Serv Res, 2007; 7:9.

Colomé

, Benseny

, Domingo

, Marí

, Caufapé

, Ferreruela

. Tele-ophthalmology eye-health system deployment in Kolda, Senegal. 2009. Available at www.academia.edu/1986642/Tele-Ophthalmology_Eye-Health_system_deployment_in_Kolda_Senegal (last accessed November 19, 2017 ).

10.

Berg

. Patient care information systems and health care work: A sociotechnical approach. Int J Med Inform, 1999; 55:87–101.

11.

Broens

, Huis in't Veld

, Vollenbroek-Hutten

, Hermens

, van Halteren

, Nieuwenhuis

. Determinants of successful telemedicine implementations: A literature study. J Telemed Telecare, 2007; 13:303–309.

12.

Godin

. Behaviors in the field of health: understanding to better act (in French). Montreal, QC: Les presses de l'université de Montréal, 2012; 20–27.

13.

Ajzen

. From intentions to actions: A theory of planed behavior. In: Kuhl

, Beckmann

, eds. Action control: From cognition to behavior. Berlin, Germany: Spinger-Verlag, 1985; 11–39.

14.

Ajzen

. The theory of planned behavior. Organ Behav Hum Decis Process, 1991; 50:179–211.

15.

Ajzen

. Attitudes, personality and behavior. Milton Keynes, England: Milton Keynes, UK Open University Press, 1988; 1–178.

16.

Kuo

, Talley

, Lee

, Yen

. The influence of telemedicine experience on physicians' perceptions regarding adoption. Telemed J E Health, 2015; 21:388–394.

17.

, Chau

. Physician acceptance of telemedicine technology: An empirical investigation. Top Health Inf Manage, 1999; 19:20–35.

18.

Hennekens

, Buring

, Mayrent

. Epidemiology in Medicine (in French). Philadelphia, PA: Lippincott-Raven Publishers, Inc., 1998; 106–110.

19.

Levin

. Study design III: Cross-sectional studies. Evid Based Dent, 2006; 7:24–25.

20.

Ancelle

. Epidemiological statistics (in French). 2ème ed. Paris, France: Maloine, 2006; 117–119.

21.

Hsieh

P-J

. Physicians'acceptance of electronic medical records exchange: An extension of the decomposed TPB model with institutional trust and perceived risk. Int J Med Inform, 2015; 84:1–14.

22.

Taylor

, Coates

. Caring from a distance: The role of telehealth. Nurs Times, 2015; 111:18–20.

23.

Hoerbst

, Schweitzer

. A systematic investigation on barriers and critical success factors for clinical information systems in integrated care settings. Yearb Med Inform, 2015; 10:79–89.

24.

Sabin

, Skimming

. A framework of ethics for telepsychiatry practice. Int Rev Psychiatry, 2015; 27:490–495.

25.

Archambault

, Bilodeau

, Gagnon

M-P

, Aubin

, Lavoie

, Lapointe

, et al. Health care professionals' beliefs about using wiki-based reminders to promote best practices in trauma care. J Med Internet Res, 2012; 14:e49.

26.

Saigí-Rubió

, Torrent-Sellens

, Jiménez-Zarco

. Drivers of telemedicine use: Comparative evidence from samples of Spanish, Colombian and Bolivian physicians. Implement Sci, 2014; 9:128.

27.

Gagnon

, Cloutier

, Fortin

. Quebec population and telehealth: A survey on knowledge and perceptions. Telemed J E Health, 2004; 10:3–12.

28.

Arne

. Ethical and legal aspects of telemedicine. Bull Acad Natl Med, 2014; 198:119–130.

29.

Dreezen

. Telemedicine and informed consent. Med Law, 2004; 23:541–549.