Abstract
Introduction:
Alcohol use disorder (AUD) is a significant cause of morbidity and mortality globally. However, uptake of AUD pharmacotherapies among clinicians has remained low. There exists a research gap regarding clinician attitudes and the diffusion of AUD pharmacotherapies among Indian clinicians. We conducted a survey to assess the same.
Methods:
An online survey was sent to clinicians nationwide from February and April 2023 to assess attitudes and practices regarding various AUD medications. Logistic and ordinal regression models evaluate the clinician and caseload factors determining medication diffusion and attitudes.
Results:
A total of 387 clinicians responded to our survey. Most of them were males (61.7%) and trained psychiatrists (56.3%). The diffusion for acamprosate, naltrexone, and disulfiram was 69.8%, 81.4%, and 58.2%, respectively. Baclofen had the highest diffusion. Female respondents had significantly lower perceived efficacy, acceptability, and safety for disulfiram. Independent practitioners were more likely to prescribe all pharmacotherapies except baclofen.
Discussion:
Diffusion for alcohol pharmacotherapies was incomplete. Baclofen had the highest diffusion, probably reflecting its cost and safety profile. Despite its perceived efficacy, safety concerns impacted the diffusion of disulfiram. Female clinicians had lower perceived attitudes toward disulfiram. Specific training for AUD pharmacotherapies is warranted for female clinicians and nonpsychiatrists.
Keywords: Alcohol pharmacotherapy, alcohol use disorder, medication attitudes, medication diffusion
Key Message:
The most commonly prescribed medications for AUD in India were Baclofen, Naltrexone, Acamprosate and Disulfiram in decreasing order. Clinician and case related variables significantly predicted attitudes towards various pharmacotherapies.
Alcohol use disorder (AUD) imposes a heavy societal, health, and financial burden. In terms of disability adjusted life years (DALYs), in 2016, the harmful use of alcohol resulted in an estimated 3 million deaths (5.3% of all deaths) and 132.6 million DALYs (5.1% of all DALYs) lost globally. This was more than other well-known causes of mortality like tuberculosis, HIV/AIDS, diabetes, hypertension, and road injuries. 1 In India, the estimated prevalence of harmful alcohol use was 5.2% (57 million people), accounting for 4.5% of the total DALYs lost.2,3 Despite the high prevalence and health impact of AUD, historically it has a large treatment gap.4,5 In India, this treatment gap for AUD is estimated to be between 75 and 83%.2,6
Over the decades, multiple pharmacotherapies have been developed to prevent relapse in AUD. 7 As of 2023, there are three United States Food and Drug Administration (USFDA) approved pharmacotherapies for AUD: disulfiram, naltrexone, and acamprosate. In addition to this, some other medicines are also used off-label, including nalmefene, gabapentin, topiramate, baclofen, and ondansetron.8,9
Prior Research on Diffusion and Physician Attitudes Toward Pharmacotherapy for AUD
Despite the availability of these medications, Western research indicates that the acceptance and uptake of these pharmacotherapies remain inconsistent and low, even among specialized addiction treatment clinicians and programs. 10 One reason for this low diffusion has been a lack of information and experience with pharmacotherapies. 11 Structural barriers like licensing requirements, personnel factors, access to physicians, and financial factors (particularly insurance coverage) have also been identified as barriers to adopting some pharmacotherapies.12,13 In some studies, a lack of acceptability for treatment has been observed not only among counsellors (who are more likely to be undertrained in pharmacotherapies) but also among clinicians.14,15
Most of the studies on clinician attitudes and diffusion of AUD pharmacotherapies have focused on naltrexone. Insurance coverage, clinicians’ perceptions of effectiveness, and exposure to information have been shown to modulate naltrexone prescriptions, which were limited by concerns about compliance and the cost of the medication.16–18 Some other variables that significantly influence pharmacotherapy use include conflict with treatment philosophy, lack of confidence in the value of pharmacotherapies, and years of experience in treating patients. 19 Clinicians’ training and background also influence medication use with general physicians prescribing USFDA-approved drugs but psychiatrists having a greater tendency to prescribe medications like antidepressants for AUD treatment. 20 Even among USFDA-approved medications, general physicians rely on disulfiram more than psychiatrists. 10 Indeed, the physician’s knowledge, familiarity, acceptability, and perceived efficacy of the drug remain indicators of the extent to which the medication has diffused in the field. This is also true for opioid use disorders. 21 Although, with time, the diffusion of some AUD pharmacotherapies (particularly naltrexone) has increased, there still exist significant attitudinal barriers toward AUD pharmacotherapies. 22
Western studies have also indicated that the adoption of pharmacotherapies among clinicians depends on factors beyond mere clinical guidelines and recommendations.11,15 Similar studies are lacking in India (or Asia). However, attitudinal differences are expected due to the widely different healthcare systems. This study investigated the attitudes and diffusion among Indian clinicians toward selected AUD pharmacotherapies. The chosen medicines include the three FDA-approved ones and three commonly used off-label medications, namely, baclofen, topiramate, and ondansetron. Multivariate ordinal models have been used to explore factors that influence clinician attitudes.
Methods
The study was conducted between February and April 2023 in a nationally representative sample of clinicians who self-reported working with clients with AUD. An online survey was designed using Google Forms https://www.google.com/forms/about/. The survey was emailed to registered members of the Addiction Psychiatry Society of India (APSI). The APSI is the largest body of clinicians and related professionals who work in the field of addiction in India. The survey was also emailed to a database of psychiatrists available with the authors (to include non-APSI members who might be treating individuals with AUD). Based on the membership of the Indian Psychiatric Society (representing almost 9,000 clinicians working in mental health), a respondent size of 350 was determined a priori to be significant.
Respondents were required to be qualified MBBS physicians registered with any medical council for at least 1 year. They self-reported having attended to at least 30 patients of AUD in the past year. A specific questionnaire was developed for the study to explore the attitudes and diffusion among respondents and their relationships to the following variables.
Independent Variables
Two general categories of independent variables were assessed: clinician-related and caseload-related. Clinician-related variables included gender, age, educational qualification, years of experience in treating AUD, and current designation. Caseload variables include locality of practice (urban vs. rural), type of health care facility, and estimated annual caseload.
Dependent Variables
A. Diffusion
The diffusion of each medication was measured as the percentage of respondents who reported having ever prescribed a medicine. Those who had never used a particular drug could respond “Never Prescribed” to a question on their use of that drug. The proportion of clinicians who had ever used that medicine was used as an index of diffusion of that particular drug.
B. Clinician Attitudes
The attitudes toward pharmacotherapies were measured using three variables: perceived effectiveness, acceptability, and safety for each medication included in the study. Respondents were asked to rate their perceived effectiveness, acceptability, or safety for each medicine on a 10-point Likert scale. Response categories ranged from 1 (not at all effective) to 10 (very effective). The clinicians were also given an option of responding “don’t know” if they felt they could not assess a medication’s efficacy, safety, or acceptability.
Analytical Strategy
A series of binary logistic regression models were estimated to predict the likelihood that clinicians reported having prescribed a medicine. Multivariate ordinal regression models examined the predictors of each medication’s perceived safety, efficacy, and acceptability. Models were estimated using the SPSS 26 software package (IBM Corp). Diagnostic tests showed no evidence of multicollinearity among independent variables.
Ethical Approval
The Research Review Board and Ethics Committee of the institute approved the research design.
Results
A total of 389 clinicians responded to the survey, yielding an estimated response rate of 15%. Respondents included clinicians possessing only the MBBS degree (MBBS doctors) and trainee psychiatrists working in government and private sectors in a nationally representative sample. The participants also included general physicians (working with AUD patients), psychiatrists working as faculty at government and private institutions, and psychiatrists involved only in clinical practice. The data was assessed for duplication and errors, and 378 entries were found valid and further analyzed.
The average age (SD) of the respondents was 33.75 (±7.3) years. Although participants were from 24 to 80 years of age, most respondents (71%) were young psychiatrists below 35 years of age, and 39.3% were females. Urban practitioners (86.2%) constituted the vast majority of respondents. In terms of educational qualification, nearly half (56.3%) had completed their master’s training and were practicing licensed psychiatrists. Notably, 27.9% were trainee psychiatrists (undergoing residency in psychiatry), and 15.8% had MBBS as their only degree. In terms of experience, the respondents had been treating AUD patients for an average of 6.86 (±6.4) years, while in terms of caseload, respondents reported treating a median number of 120 (IQR 60-250) cases of AUD per year.
Table 1 provides information about the diffusion, perceived efficacy, safety, and acceptability of various pharmacotherapies among clinicians treating AUD. The diffusion values of acamprosate, naltrexone, and disulfiram were 69.8%, 81.4%, and 58.2%, respectively, while those of baclofen, topiramate, and ondansetron were 85.7%, 65.3%, and 34.9%. As a group, the diffusion for the FDA-approved group was higher than the non-FDA-approved medicines but only slightly.
Table 1.
Diffusion, Perceived Efficacy, Safety, and Acceptability of Clinicians Toward Alcohol Pharmacotherapies.
| Dependent Variables | FDA-approved Pharmacotherapies | Non-FDA-approved Pharmacotherapies | ||||
| Acamprosate | Naltrexone | Disulfiram | Topiramate | Baclofen | Ondansetron | |
| Diffusiona | 69.8 | 81.4 | 58.2 | 65.3 | 85.7 | 34.9 |
| Perceived Efficacyb | 8(6–9) | 8(6–9) | 8(7–9) | 5(3–7) | 7(5–8) | 4(3–6) |
| Perceived Safetyb | 8(7–9) | 8(6–9) | 4(2–7) | 7(5–9) | 9(7–10) | 8(6–9) |
| Perceived Acceptabilityb | 7(5–8) | 8(6–10) | 5(3–7) | 5(3–7) | 9(7–10) | 5(3–7) |
aPercentage of clinicians who reported having prescribed each pharmacotherapy, b Item measured on a scale of 1–10, median (inter-quartile range) reported.
Regarding perceived scores, the median perceived efficacy for acamprosate, naltrexone, and disulfiram was the highest, followed by baclofen, topiramate, and ondansetron. Disulfiram had the lowest median perceived safety. Baclofen was perceived to be the safest drug, followed by acamprosate, naltrexone, and ondansetron, with the same median safety rating. Baclofen also had the highest median acceptability, followed by naltrexone and acamprosate. The median perceived acceptability of disulfiram, topiramate, and ondansetron were the same and the lowest among all pharmacotherapies.
Modeling the Diffusion of Alcohol Pharmacotherapies
Binary logistic regression models predicting the probability of having prescribed each medication are presented in Table 2. Age and locality of practice were the two clinician characteristics that were not significant factors in predicting the diffusion of any drug. Independent practitioners were found to be significantly more likely to have ever prescribed all pharmacotherapies, namely, acamprosate (OR 2.57, 95% CI: 1.21–5.46, p<0.05), naltrexone (OR 2.53, 95% CI: 1.18–5.46 p < 0.05), disulfiram (OR 48.85, 95% CI: 14.44–165.25, p<0.01), and topiramate (OR: 2.63, 95% CI: 1.11–6.24, p<0.05), except baclofen for which a significance could not be established. Independent practitioners were less likely to have ever used ondansetron (OR: 0.06, 95% CI: 0.01–0.05, p <0.05). Female respondents were significantly less likely to have prescribed disulfiram (OR 0.04, CI: 0.01–0.10, p<0.01). Residency trainees were significantly more likely to have prescribed naltrexone (OR 3.2, CI: 6.67–0.65, p <0.01) and disulfiram (OR 52.54, CI: 16.03–172.18, p<0.01) but were significantly less likely to have prescribed ondansetron (OR 0.01, CI: 0.00 –0.01, p <0.01).
Table 2.
Binary Logistic Regression for Diffusion of Alcohol Pharmacotherapies.
| Predictors | Acamprosate | Naltrexone | Disulfiram | Topiramate | Baclofen | Ondansetron | ||||||
| b (SE) | OR (95% CI) |
b (SE) | OR (95% CI) |
b (SE) | OR (95% CI) |
b (SE) | OR (95% CI) |
b (SE) | OR (95% CI) |
b (SE) | OR (95% CI) |
|
| AGE | –0.060 (0.061) |
0.94 (0.83–1.06) |
0.010 (0.046) |
1.01 (0.92–1.10) |
0.036 (0.053) |
1.04 (0.93–1.15) |
0.018 (0.064) |
1.02 (0.89–1.15) |
–0.830 (0.425) |
0.44 (0.18–1.03) |
0.031 (0.064) |
1.03 (0.91–1.17) |
| EXPERIENCE |
0.516
†
(0.103) |
1.67 (1.36–2.05) |
–0.087 (0.054) |
0.91 (0.82–1.01) |
–0.072 (0.064) |
0.93 (0.82–1.05) |
0.036 (0.069) |
0.91 (0.90–1.18) |
0.391 (0.548) |
0.47 (0.50–4.32) |
0.053 (0.069) |
1.05 (0.92–1.21) |
| ANNUAL CASELOAD | –0.002* (0.001) |
1.02 (1.02–1.04) |
–0.001 (0.001) |
0.99 (0.99–1.00) |
0.001 (0.001) |
0.99 (0.99–1.00) |
0.001 (0.001) |
1.00 (0.99–1.00) |
0.003 (0.002) |
1.03 (0.99–1.01) |
0.003
†
(0.001) |
1.03 (1.00–1.05) |
| FEMALE | –0.192 (0.304) |
0.82 (0.46–1.49) |
–0.209 (0.314) |
0.81 (0.43–1.50) |
3.257
†
(0.490) |
0.04 (0.01–0.10) |
–0.291 (0.321) |
0.39 (0.39–1.40) |
–1.238 (0.877) |
0.29 (0.05–1.68) |
0.023 (0.375) |
1.02 (0.49–2.13) |
| COMPLETED RESIDENCY | –0.085 (0.383) |
0.92 (0.43–1.95) |
–1.164† (0.375) |
3.20 (6.67–0.65) |
–3.962† (0.606) |
52.54 (16.03–172.18) |
–0.332 (0.399) |
0.72 (0.32–1.56) |
–22.489 (1749.670) |
0.86 (0.62–1.24) |
–6.677† (1.116) |
0.01 (0.00–0.01) |
| INDEPENDENT PRACTIONER |
0.944*
(0.385) |
2.57 (1.21–5.46) |
0.932*
(0.391) |
2.53 (1.18–5.46) |
–3.889† (0.622) |
48.85 (14.44–165.25) |
0.968*
(0.441) |
2.63 (1.11–6.24) |
–28.648 (1906.353) |
1.22 (0.84–1.46) |
–5.128* (1.083) |
0.06 (0.01–0.05) |
| FACULTY | –19.754 (34.771) |
0.12 (0.12–1.24) |
–0.657 (0.590) |
0.51 (0.16–1.65) |
0.572 (0.559) |
1.77 (0.59–5.29) |
–0.098 (0.481) |
0.91 (0.35–2.32) |
–22.242 (2500.819) |
1.06 (0.42–1.28) |
0.350 (0.479) |
1.42 (0.55–3.63) |
| PRIVATE INSTITUTION | 0.313 (0.303) |
1.37 (0.75–2.47) |
0.176 (0.298) |
1.19 (0.66–2.14) |
4.184
†
(0.527) |
23.38 (23.38–184.13) |
–21.918 (29.149) |
0.68 (0.45–1.36) |
–22.631 (2095.739) |
1.12 (0.84–1.88) |
–0.199 (0.355) |
0.82 (0.41–1.64) |
| URBAN | –0.227 (0.468) |
0.78 (0.32–1.99) |
–0.741 (0.383) |
0.48 (0.22–1.01) |
–0.529 (0.469) |
0.59 (0.23–1.47) |
–0.132 (0.482) |
0.87 (0.34–2.25) |
0.417 (1.292) |
1.52 (0.12–19.09) |
0.133 (0.499) |
1.14 (0.42–3.04) |
| Cox & Snell R 2 | 0.396 | 0.445 | 0.538 | 0.46 | 0.512 | 0.47 | ||||||
* p < 0.05, †p < 0.01, CI = confidence interval, *p < 0.05; †p < 0.01 (two-tailed significance test).
Among medications, for acamprosate, years of experience (OR 1.67, CI: 1.36–2.05, p <0.01) and annual caseload (OR 1.02, CI: 1.02–1.04, p<0.05) were significant factors in its diffusion. Annual caseload was also a significant predictor of diffusion of ondansetron (OR 1.03, CI: 1.00–1.05, p<0.01). Clinicians working in a privately owned clinic/institution were significantly more likely to have prescribed disulfiram (OR 23.38, CI: 23.38–184.13, p<0.05).
Modeling of Perceived Effectiveness of Alcohol Pharmacotherapies
An ordinal regression model was used to examine the clinician variables that predicted the perceived effectiveness of various AUD pharmacotherapies, the results of which are summarized in Table 3. Age significantly predicted (but negatively) respondents’ perceived efficacy of disulfiram (95% CI; –0.041 to 0.008, p<0.05) and acamprosate (95% CI: –0.119 to 0.004, p<0.05). Experience was a significant predictor of favorable perceived efficacy among respondents for acamprosate (95% CI: 0.056–0.084, p<0.05) and topiramate (95% CI: 0.045–0.093, p<0.01). Annual caseload significantly predicted favorable perception in efficacy for disulfiram (95% CI: 0.00–0.002, p<0.05) and baclofen (95% CI: 0.001–0.012, p<0.05). The female respondents were significantly more likely to have a reduced perception of disulfiram’s efficacy (95% CI: –0.747 to –0.002, p<0.05). Faculty were significantly more likely to have favorable perceived efficacy for naltrexone (95% CI: 0.073–1.111, p<0.05) and topiramate (CI: 0.384–1.557, p<0.01). In comparison, the independent practitioners were significantly more likely to favorably rate efficacy for disulfiram (95% CI: 0.116–1.115, p<0.05), naltrexone (95% CI: 0.30 – 0.711, p<0.05), and topiramate (95% CI: –0.463 to 0.529, p<0.01). Clinicians working in a private institution were significantly more likely to favorably rate the efficacy of topiramate (95% CI: 0.518–1.251, p<0.01) and baclofen (95% CI: –0.307 to –0.415, p<0.05).
Table 3.
Ordinal Regression for Perceived Efficacy of Alcohol Pharmacotherapies.
| Predictors | Acamprosate | Naltrexone | Disulfiram | Topiramate | Baclofen | Ondansetron | ||||||
| Estimate (SE) | 95% CI | Estimate (SE) | 95% CI | Estimate (SE) | 95% CI | Estimate (SE) | 95% CI | Estimate (SE) | 95% CI | Estimate (SE) | 95% CI | |
| AGE | –0.057* (0.031) |
–0.119 –0.004 |
–0.047 (0.031) |
–0.109 0.014 |
–0.019* (0.031) |
–0.041 –0.008 |
–0.044 (0.031) |
–0.104 0.017 |
–0.031 (0.031) |
–0.092 0.030 |
0.002 (0.031) |
–0.058 0.063 |
| EXPERIENCE |
0.074*
(0.036) |
0.056 0.084 |
0.022 (0.036) |
–0.048 0.092 |
0.034 (0.035) |
0.035 0.103 |
0.064
†
(0.035) |
0.045 0.093 |
0.023 (0.035) |
–0.047 0.092 |
0.007 (0.035) |
–0.062 0.076 |
| ANNUAL CASELOAD | 0.004 (0.001) |
–0.008 0.002 |
0.001 (0.001) |
0.000 0.002 |
0.001*
(0.001) |
0.0003951 0.002 |
0.001 (0.001) |
0.000 0.002 |
0.007*
(0.001) |
0.001 0.012 |
0.001 (0.001) |
–0.001 0.002 |
| FEMALE | 0.049 (0.194) |
–0.331 0.429 |
0.130 (0.194) |
–0.251 0.510 |
–0.375* (0.190) |
–0.747 –0.002 |
0.041 (0.189) |
–0.330 0.411 |
0.117 (0.191) |
–0.258 0.491 |
–0.257 (0.190) |
–0.629 0.114 |
| COMPLETED RESIDENCY | 0.154 (0.241) |
–0.318 0.626 |
0.129 (0.241) |
–0.344 0.602 |
–0.114 (0.237) |
–0.579 0.351 |
–0.082 (0.237) |
–0.546 0.382 |
–0.074 (0.240) |
–0.544 0.396 |
–0.111 (0.237) |
–0.576 0.354 |
| INDEPENDENT PRACTIONER |
0.179
(0.257) |
–0.325 0.684 |
0.205*
(0.258) |
0.300 0.711 |
0.615*
(0.255) |
0.116 1.115 |
0.033
†
(0.253) |
–0.463 0.529 |
–0.273 (0.257) |
–0.777 0.231 |
–0.301 (0.254) |
–0.799 0.197 |
| FACULTY |
0.753*
(0.303) |
0.159 1.347 |
0.519*
(0.302) |
0.073 1.111 |
0.245
(0.296) |
–0.336 0.825 |
0.970† (0.299) |
0.384 1.557 |
0.240 (0.299) |
–0.347 0.826 |
0.328 (0.296) |
–0.252 0.908 |
| PRIVATE INSTITUTION | –0.222 (0.186) |
–0.588 0.143 |
0.060 (0.186) |
–0.305 0.426 |
0.349 (0.183) |
–0.010 0.708 |
0.884
†
(0.187) |
0.518 1.251 |
–0.364† (0.184) |
–0.307 –0.415 |
0.112 (0.183) |
–0.246 0.470 |
| URBAN | 0.281 (0.274) |
–0.256 0.819 |
0.462 (0.275) |
–0.076 1.000 |
–0.219 (0.271) |
–0.750 0.312 |
0.498 (0.271) |
–0.033 1.028 |
–0.023 (0.271) |
–0.555 0.508 |
0.207 (0.271) |
–0.324 0.737 |
| Cox & Snell R 2 | 0.55 | 0.38 | 0.68 | 0.26 | 0.24 | 0.34 | ||||||
* p < 0.05, †p < 0.01, CI = confidence interval, *p < 0.05; †p < 0.01 (two-tailed significance test).
Perceived Safety of Alcohol Pharmacotherapies
Similarly, Table 4 shows the results of similar models measuring the perceived safety of the six AUD pharmacotherapies. Increasing age negatively predicted the perceived safety of acamprosate (95% CI: -0.194 to -0.069, p<0.01) and baclofen (95% CI: -0.135 to -0.012, p<0.05), suggesting younger clinicians were significantly more likely to perceive them as safe. Experience significantly predicted (positively) the perceived safety of acamprosate (95% CI: 0.056–0.199, p<0.01) and ondansetron (95% CI: –0.038 to 0.101, p<0.05), while annual caseload significantly predicted an increased safety rating for baclofen (95% CI: 0.00 –0.003, p<0.05). Female clinicians were significantly more likely to perceive naltrexone as safe (95% CI: 0.754 to –0.004, p<0.05) and significantly more likely to perceive disulfiram as less safe (95% CI: –0.810 to –0.062, p<0.05). Clinicians who had completed post-graduate training in psychiatry were significantly more likely to find disulfiram (95% CI: -1.131 to –0.190, p<0.01), topiramate (95% CI: -1.068 to –0.133, p<0.05), and ondansetron (95% CI: –0.937 to –0.001, p<0.05) less safe. Independent practitioners were significantly more likely to rate naltrexone (95% CI: 0.958–1.995, p<0.01), disulfiram (95% CI: 0.035–1.205, p<0.05), and baclofen (95% CI: 0.344–1.371, p<0.01) as safe. Similarly, faculty were significantly more likely to rate all medications favorably safe than disulfiram (acamprosate 95% CI: 0.797–2.008, p<0.01; naltrexone 95% CI: 1.033–2.239, p<0.01; topiramate 95% CI: 0.302–1.472, p<0.01; baclofen 95% CI: 0.564–1.764; and ondansetron 95% CI: 0.062–1.232, p <0.05). Clinicians working in a private institution were significantly more likely to have a favorable view of the safety of naltrexone (95% CI: 1.470–2.271, p<0.05). Finally, clinicians working in an urban environment were significantly more likely to have an increased perception of the safety of naltrexone (95% CI: 0.112–1.185, p<0.01) and baclofen (95% CI: -1.271 to –0.157, p<0.05).
Table 4.
Ordinal Regression for the Perceived Safety of Alcohol Pharmacotherapies.
| Predictors | Acamprosate | Naltrexone | Disulfiram | Topiramate | Baclofen | Ondansetron | ||||||
| Estimate (SE) | 95% CI | Estimate (SE) | 95% CI | Estimate (SE) | 95% CI | Estimate (SE) | 95% CI | Estimate (SE) | 95% CI | Estimate (SE) | 95% CI | |
| AGE | –0.132† (0.032) |
–0.194 –0.069 |
–0.051 (0.031) |
–0.112 0.011 |
0.001 (0.031) |
–0.059 0.062 |
–0.056 (0.031) |
–0.117 0.005 |
–0.073* (0.031) |
–0.135 –0.012 |
–0.028 (0.031) |
–0.089 0.033 |
| EXPERIENCE |
0.127
†
(0.036) |
0.056 0.199 |
0.030 (0.036) |
–0.040 0.099 |
0.018 (0.035) |
–0.051 0.088 |
0.052 (0.035) |
–0.018 0.121 |
0.037 (0.036) |
–0.033 0.107 |
0.031*
(0.035) |
–0.038 0.101 |
| ANNUAL CASELOAD | 0.001 (0.001) |
0.000 0.002 |
0.000 (0.001) |
–0.001 0.002 |
0.001 (0.001) |
0.000 0.002 |
0.000 (0.001) |
–0.001 0.002 |
0.002*
(0.001) |
0.000 0.003 |
0.002 (0.001) |
0.000 0.003 |
| FEMALE | –0.309 (0.193) |
–0.688 0.070 |
–0.379* (0.192) |
0.754 –0.004 |
–0.436* (0.191) |
–0.810 –0.062 |
–0.063 (0.189) |
–0.433 0.308 |
–0.206 (0.193) |
–0.584 0.173 |
–0.195 (0.190) |
–0.567 0.178 |
| COMPLETED RESIDENCY | –0.046 (0.240) |
–0.517 0.425 |
–0.015 (0.239) |
–0.483 0.454 |
–0.660† (0.240) |
–1.131 –0.190 |
–0.601* (0.239) |
–1.068 –0.133 |
–0.470 (0.243) |
–0.945 0.006 |
–0.468* (0.239) |
–0.937 –0.001 |
| INDEPENDENT PRACTIONER | 0.023 (0.257) |
–0.480 0.527 |
1.476
†
(0.265) |
0.958 1.995 |
0.620*
(0.298) |
0.035 1.205 |
0.432 (0.254) |
–0.067 0.930 |
0.857
†
(0.262) |
0.344 1.371 |
0.190 (0.255) |
–0.309 0.689 |
| FACULTY |
1.403
†
(0.309) |
0.797 2.008 |
1.636
†
(0.308) |
1.033 2.239 |
0.371 (0.255) |
–0.129 0.871 |
0.887
†
(0.299) |
0.302 1.472 |
1.164
†
(0.306) |
0.564 1.764 |
0.647*
(0.299) |
0.062 1.232 |
| PRIVATE INSTITUTION | –0.001 (0.186) |
–0.366 0.365 |
1.871
†
(0.205) |
1.470 2.271 |
0.305 (0.183) |
–0.054 0.665 |
–0.052 (0.182) |
–0.410 0.305 |
–0.034 (0.186) |
–0.400 0.331 |
0.100 (0.184) |
–0.260 0.460 |
| URBAN | –0.115 (0.277) |
–0.657 0.428 |
0.648*
(0.274) |
0.112 1.185 |
0.261 (0.271) |
–0.270 0.793 |
0.191 (0.270) |
–0.338 0.720 |
–0.714* (0.284) |
–1.271 –0.157 |
–0.294 (0.273) |
–0.829 0.241 |
| Cox & Snell R 2 | 0.146 | 0.308 | 0.434 | 0.382 | 0.472 | 0.467 | ||||||
* p < 0.05, †p < 0.01, CI = confidence interval, *p < 0.05; †p < 0.01 (two-tailed significance test).
Perceived Acceptability of Alcohol Pharmacotherapies
Table 5 displays the results of ordinal regression models for clinicians’ perception of the acceptability of AUD pharmacotherapies. Significance was established for age and experience. Increasing age reduced the perceived acceptability for acamprosate (95% CI: –0.146 to 0.024, p<0.01), disulfiram (95% CI: –0.134 to 0.013, p<0.05), and baclofen (95% CI: –0.106 to 0.015, p<0.05). Conversely, increasing experience significantly predicted increased acceptability of acamprosate (95% CI: 0.053–0.194, p<0.01), disulfiram (95% CI: 0.008–0.146, p<0.05), topiramate (95% CI: 0.003–0.142, p<0.05), and baclofen (95% CI: 0.052–0.086, p <0.05). Annual caseload significantly predicted increased acceptability of only ondansetron (95% CI: 0.001–0.032, p<0.05). As with perceived acceptability and safety, female respondents were more likely to have reduced acceptability of disulfiram (95% CI: –0.240 to –0.500, p<0.01). Completing postgraduate training in psychiatry was unlikely to significantly predict acceptability toward any medication. Faculty were significantly more likely to favorably regard the acceptability of acamprosate (95% CI: 0.822–2.017, p <0.01), naltrexone (95% CI: 1.321–2.539, p <0.01), topiramate (95% CI: 1.351–2.567, p <0.05), and ondansetron (95% CI: 0.060–1.105, p <0.05). Those with an independent practice had significantly more acceptability of naltrexone (95% CI: 0.004–1.004, p<0.05) and disulfiram (95% CI: 0.075–0.918, p<0.01). Similarly, practitioners working in a private institution had significantly greater acceptability of naltrexone (95% CI: 0.617–1.357, p<0.01) and disulfiram (95% CI: 0.086–0. 803, p<0.01). Finally, an urban or a rural area of practice did not predict the perceived acceptability of any of the medications.
Table 5.
Ordinal Regression for Perceived Acceptability of Alcohol Pharmacotherapies.
| Predictors | Acamprosate | Naltrexone | Disulfiram | Topiramate | Baclofen | Ondansetron | ||||||
| Estimate (SE) | 95% CI | Estimate (SE) | 95% CI | Estimate (SE) | 95% CI | Estimate (SE) | 95% CI | Estimate (SE) | 95% CI | Estimate (SE) | 95% CI | |
| AGE | –0.085† (0.031) |
–0.146 –0.024 |
–0.017 (0.031) |
–0.077 0.044 |
–0.073* (0.031) |
–0.134 –0.013 |
–0.032 (0.031) |
–0.092 0.029 |
–0.046* (0.031) |
–0.106 –0.015 |
–0.003 (0.031) |
–0.063 0.058 |
| EXPERIENCE |
0.123
†
(0.036) |
0.053 0.194 |
0.026 (0.035) |
–0.043 0.096 |
0.077*
(0.035) |
0.008 0.146 |
0.073*
(0.035) |
0.003 0.142 |
0.072*
(0.035) |
0.052 0.086 |
0.031 (0.035) |
–0.039 0.100 |
| ANNUAL CASELOAD | –0.001 (0.001) |
–0.002 0.000 |
0.000 (0.001) |
–0.002 0.001 |
0.000 (0.001) |
–0.001 0.002 |
0.001 (0.001) |
–0.001 0.002 |
0.001 (0.001) |
–0.001 0.002 |
0.018*
(0.001) |
0.001 0.032 |
| FEMALE | –0.171 (0.190) |
–0.543 0.202 |
–0.013 (0.190) |
–0.385 0.360 |
–0.130† (0.189) |
–0.240 –0.500 |
0.121 (0.189) |
–0.250 0.491 |
–0.042 (0.190) |
–0.414 0.330 |
0.114 (0.190) |
–0.259 0.486 |
| COMPLETED RESIDENCY | –0.050 (0.238) |
–0.516 0.415 |
–0.074 (0.238) |
–0.540 0.391 |
0.027 (0.236) |
–0.436 0.491 |
0.186 (0.237) |
–0.278 0.650 |
–0.069 (0.238) |
–0.535 0.397 |
0.204 (0.238) |
–0.263 0.671 |
| INDEPENDENT PRACTIONER | 0.134 (0.254) |
–0.363 0.632 |
0.504* (0.255) |
0.004 1.004 |
0.422
†
(0.253) |
0.075 0.918 |
0.461 (0.254) |
–0.037 0.960 |
0.129 (0.254) |
–0.370 0.627 |
–0.201 (0.255) |
–0.701 0.299 |
| FACULTY |
1.420
†
(0.305) |
0.822 2.017 |
1.930
†
(0.311) |
1.321 2.539 |
0.067 (0.295) |
–0.511 0.644 |
1.959*
(0.310) |
1.351 2.567 |
0.141 (0.297) |
–0.440 0.723 |
0.523* (0.297) |
0.060 1.105 |
| PRIVATE INSTITUTION | –0.230 (0.184) |
–0.590 0.130 |
0.987
†
(0.189) |
0.617 1.357 |
0.444
†
(0.183) |
0.086 0.803 |
0.138 (0.183) |
–0.220 0.496 |
–0.175 (0.183) |
–0.534 0.184 |
–0.124 (0.183) |
–0.483 0.235 |
| URBAN | 0.039 (0.271) |
–0.493 0.570 |
–0.291 (0.272) |
–0.823 0.242 |
–0.307 (0.270) |
–0.835 0.222 |
–0.110 (0.270) |
–0.639 0.419 |
0.275 (0.270) |
–0.255 0.805 |
0.063 (0.271) |
–0.470 0.595 |
| Cox & Snell R 2 | 0.144 | 0.217 | 0.448 | 0.241 | 0.245 | 0.572 | ||||||
* p < 0.05, †p < 0.01, CI = confidence interval, *p < 0.05; †p < 0.01 (two-tailed significance test).
Discussion
This study explored the diffusion of alcohol pharmacotherapies among Indian clinicians and their perceived attitudes toward various pharmacotherapies. Although similar studies have been done in Western countries, 11 to the best of our knowledge, this is the first study in India, and some of the findings in the study suggest some differences with Western data. Respondents in the study were a young cohort with a large number of female respondents, and the results of the study are more likely to indicate the attitudes of the next generation of addiction psychiatrists in India.
Baclofen (a non-FDA-approved medication) had the highest diffusion, followed by naltrexone. This was surprising since the evidence for the efficacy of baclofen in AUD is far from clinching. The diffusion of the remaining medicines was similar to Western literature, with acamprosate and disulfiram following naltrexone in diffusion among AUD clinicians. 22 Although higher than their Western counterparts, the diffusion of many pharmacotherapies is far from complete among Indian clinicians. A sizeable percentage of clinicians reported having never prescribed disulfiram despite being the oldest available and approved medication for AUD. In fact, more clinicians reported having prescribed topiramate (another non-FDA-approved medication) than disulfiram. This may reflect clinician concerns about the adverse effects of disulfiram. It may also reflect the need to specially train clinicians that supervised disulfiram is both safe and efficacious. 23 The diffusion of ondansetron was also higher than expected, though the lowest among all the studied medications. The reporting of ever using a medicine was calculated as diffusion of that pharmacotherapy. In this context, the diffusion of ondansetron was more reflective of clinicians having ever used it rather than prescription rates.
The diffusion of naltrexone was greater than in previous studies and even greater than in a recent study.12,22,24,25 This may reflect progress in the diffusion of naltrexone or differences in clinical practices compared to Western nations. In fact, the diffusion of all medications was higher in this study than in previous research and probably reflects the greater emphasis on pharmacotherapy among Indian clinicians.10,26
The results of the attitude questions suggest that specific training may be required among certain groups to enhance the prescription of specific pharmacotherapies. Although the perceived efficacy of disulfiram was comparable to acamprosate and naltrexone, lower perceived safety was reflected in the reduced acceptability of disulfiram. As with diffusion, this probably indicates the need for specific training regarding the rates of adverse effects with disulfiram are comparable to other medications. 27
As a group, female clinicians were less likely to rate disulfiram as efficacious, safe, and acceptable. Such beliefs among specific clinician groups are not new, and researchers have devised several strategies to address informational, provider, and patient level barriers to adopting pharmacotherapies.28,29 Such strategies may also be adopted in the Indian context, especially as part of continued medical education. Training should also be tailored to meet the specific needs of the target population (e.g., psychiatrists vs. general physicians) and use flexible options (e.g., web-based deliveries) and repeated to increase participation and medicine uptake. 30 It was also observed that the diffusion of the naltrexone and disulfiram was less in those who were not qualified psychiatrists. Since this group included those whose qualifications were only MBBS, it can indicate that MBBS-qualified physicians were not prescribing these medications. This highlights the unique training needs of these physicians regarding AUD pharmacotherapy.
Training, by itself, will not be sufficient in scenarios where other factors determine clinician attitudes. In our study, independent practitioners were more likely to have a positive attitude toward most pharmacotherapies. Independent practitioners were more likely to have been working in private clinics where patients were more likely to afford pharmacotherapies. The cost of medications has been identified as a barrier to the adoption of pharmacotherapies. 17 Insurance for mental health care (or rather the lack of it) has been a topic of some concern. 31 It is hoped that recent steps by the Insurance Regulatory and Development Authority of India to increase mental health coverage also involve substance use treatment.32,33 Such a step would also be expected to increase the adoption of pharmacotherapies.
One particular advantage of the study was that it explored attitudes toward some commonly used off-label pharmacotherapies. This helped us determine the preference of some clinician demographics for baclofen. Increasing clinical experience (age, years of experience, and annual caseload) resulted in a more favorable attitude toward baclofen’s efficacy, safety, and acceptability. The preference of Indian clinicians toward baclofen was surprising and might be reflective of a difference in clinical practice or specific factors unique to India.
Overall, the results of the study indicate that there exists significant variation in the diffusion of AUD pharmacotherapies. The diffusion of FDA-approved pharmacotherapies correlated with Western scientific evidence and was higher than in Western nations. However, the diffusion of non-FDA-approved pharmacotherapies (especially baclofen) was greater than expected. Certain clinician characteristics predict clinician attitudes to specific pharmacotherapies. These characteristics may be used to design interventions for specific clinician groups to increase the uptake of pharmacotherapies.
Further qualitative studies among clinicians will be required to qualify the barriers associated with specific pharmacotherapies.
Limitations
The study was an exclusively online study. Although due care was used to distribute the response link to clinicians working with AUD clients, and the data was manually crosschecked to remove duplicate and erroneous entries, some of the limitations and biases associated with online surveys remained. We collected data only about the clinician’s attitudes and their prescribing patterns. However, patient profile and medicine availability in the hospital are two among other factors that can also influence medication diffusion and such information could have supplanted the study data.
Conclusions
Regardless, the study results indicate important aspects of the diffusion and attitudes of clinicians toward AUD pharmacotherapies in India. Specifically, the study explored diffusion and clinician attitudes toward non-FDA pharmacotherapies, a topic that has not been previously explored before.
Footnotes
The authors have no potential conflicts of interest to declare with respect to the research, authorship, and/or publication of this article.
Declaration Regarding the Use of Generative AI: None used.
Ethical Approval: This study was approved by the Institutional Human Ethics Committee, 3rd Floor Academic Block, Department of Psychiatry, All India Institute of Medical Sciences, Saket Nagar, Bhopal 462043, India. Approval Number: IHEC-LOP/2023/IL055
Funding: The authors received no financial support for the research, authorship, and/or publication of this article.
References
- 1.World Health Organization. Global status report on alcohol and health 2018. Geneva: World Health Organization, 2019. [Google Scholar]
- 2.Ambekar A, Agrawal A, Rao R, et al. Magnitude of substance use in India. New Delhi: Ministry of Social Justice and Empowerment, Government of India. [Google Scholar]
- 3.Rehm J, Mathers C, Popova S, et al. Global burden of disease and injury and economic cost attributable to alcohol use and alcohol-use disorders. Lancet, 2009; 373: 2223–2233. [DOI] [PubMed] [Google Scholar]
- 4.Hasin DS, Stinson FS, Ogburn E, et al. Prevalence, correlates, disability, and comorbidity of DSM-IV alcohol abuse and dependence in the United States: Results from the National Epidemiologic Survey on alcohol and related conditions. Arch Gen Psychiatry, 2007; 64: 830–842. [DOI] [PubMed] [Google Scholar]
- 5.Substance Abuse and Mental Health Services Administration. Behavioral health trends in the United States: Results from the 2014 National Survey on Drug Use and Health 2015 HHS Publication No. SMA 15-4927, NSDUH Series H-50. Rockville, MD: Substance Abuse and Mental Health Services Administration. [Google Scholar]
- 6.Gautham MS, Gururaj G, Varghese M, et al. The National Mental Health Survey of India (2016): Prevalence, socio-demographic correlates and treatment gap of mental morbidity. Int J Soc Psychiatry, 2020; 66: 361–372. [DOI] [PubMed] [Google Scholar]
- 7.Adinoff B. Neurobiologic processes in drug reward and addiction. Harv Rev Psychiatry, 2004; 12: 305–320. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Swift RM and Aston ER. Pharmacotherapy for alcohol use disorder: Current and emerging therapies. Harv Rev Psychiatry, 2015; 23: 122–133. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Kranzler HR and Soyka M.. Diagnosis and pharmacotherapy of alcohol use disorder: A review. JAMA, 2018; 320: 815–824. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Ducharme LJ, Knudsen HK and Roman PM. Trends in the adoption of medications for alcohol dependence. J Clin Psychopharmacol, 2006; 26: S13–S19. [DOI] [PubMed] [Google Scholar]
- 11.Abraham AJ, Ducharme LJ and Roman PM. Counselor attitudes toward pharmacotherapies for alcohol dependence. J Stud Alcohol Drugs, 2009; 70: 628–635. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Thomas CP, Wallack SS, Lee S, et al. Research to practice: Adoption of naltrexone in alcoholism treatment. J Subst Abuse Treatment, 2003; 24: 1–11. [PubMed] [Google Scholar]
- 13.Oser CB and Roman PM. Organizational-level predictors of adoption across time: Naltrexone in private substance-use disorders treatment centers. J Stud Alcohol Drugs, 2007; 68: 852–861. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Harris AHS, Oliva E, Bowe T, et al. Pharmacotherapy of alcohol use disorders by the veterans health administration: Patterns of receipt and persistence. Psychiatr Serv, 2012; 63: 679–685. [DOI] [PubMed] [Google Scholar]
- 15.Meza EE, Cunningham JA, el-Guebaly N, et al. Alcoholism: Beliefs and attitudes among Canadian alcoholism treatment practitioners. Can J Psychiatry, 2001; 46: 167–172. [DOI] [PubMed] [Google Scholar]
- 16.Mark TL, Kranzler HR, Song X, et al. Physicians’ opinions about medications to treat alcoholism. Addiction, 2003; 98: 617–626. [DOI] [PubMed] [Google Scholar]
- 17.Mark TL, Kranzler HR and Song X.. Understanding US addiction physicians’ low rate of naltrexone prescription. Drug Alcohol Dependence, 2003; 71: 219–228. [DOI] [PubMed] [Google Scholar]
- 18.Mark TL, Kranzler HR, Poole VH, et al. Barriers to the use of medications to treat alcoholism. Am J Addict, 2003; 12: 281–294. [PubMed] [Google Scholar]
- 19.Thomas SE and Miller PM. Knowledge and attitudes about pharmacotherapy for alcoholism: A survey of counselors and administrators in community-based addiction treatment centres. Alcohol Alcohol, 2007; 42: 113–118. [DOI] [PubMed] [Google Scholar]
- 20.Ponce Martinez C, Vakkalanka P and Ait-Daoud N. Pharmacotherapy for alcohol use disorders: Physicians’ perceptions and practices. Front Psychiatry, 2016; 7: 182. doi: https://doi.org/10.3389/fpsyt.2016.00182 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Mark TL, Kassed CA, Vandivort-Warren R, et al. Alcohol and opioid dependence medications: Prescription trends, overall and by physician specialty. Drug Alcohol Depend, 2009; 99: 345–349. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Ehrie J, Hartwell EE, Morris PE, et al. Survey of addiction specialists’ use of medications to treat alcohol use disorder. Front Psychiatry, 2020; 11: 47. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Mutschler J, Dirican G, Gutzeit A, et al. Safety and efficacy of long-term disulfiram aftercare. Clin Neuropharmacol, 2011; 34: 195. [DOI] [PubMed] [Google Scholar]
- 24.Petrakis IL, Leslie D and Rosenheck R.. Use of naltrexone in the treatment of alcoholism nationally in the Department of Veterans Affairs. Alcohol Clin Exp Res, 2003; 27: 1780–1784. [DOI] [PubMed] [Google Scholar]
- 25.Fuller BE, Rieckmann T, McCarty D, et al. Adoption of naltrexone to treat alcohol dependence. J Subst Abuse Treatm, 2005; 28: 273–280. [DOI] [PubMed] [Google Scholar]
- 26.Knudsen HK, Abraham AJ and Roman PM. Adoption and implementation of medications in addiction treatment programs. J Addict Med, 2011; 5: 21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Bhatia G, Sarkar S, Adagadda SS, et al. Disulfiram safety in alcohol use disorders: Experience from an addiction treatment center in India. Indian J Psychiatry, 2022; 64: 216–217. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Oliva EM, Maisel NC, Gordon AJ, et al. Barriers to use of pharmacotherapy for addiction disorders and how to overcome them. Curr Psychiatry Rep, 2011; 13: 374–381. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Thomas SE, Miller PM, Randall PK, et al. Improving acceptance of naltrexone in community addiction treatment centers: A pilot study. J Subst Abuse Treatm, 2008; 35: 260–268. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Fitzgerald J and McCarty D.. Understanding attitudes toward use of medication in substance abuse treatment: A multilevel approach. Psychol Serv, 2009; 6: 74. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Ghosh M and Chatterji BP. Mental health insurance in India after COVID-19. Lancet Psychiatry, 2020; 7: 1016–1017. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Insurance Regulatory and Development Authority. Providing cover for Mental Illness under Health Insurance policies. Hyderabad: Insurance Regulatory and Development Authority, 2022. [Google Scholar]
- 33.Insurance Regulatory and Development Authority. Expert committee for maters related to mental health. https://irdai.gov.in/web/guest/document-detail?documentId=3388900 (2023, accessed 29 June 2023. ).