Patient attitudes to and satisfaction with their treatment are associated with improved adherence. There is a paucity of data on patient drug attitudes and preference to oral compared to long-acting injectable (LAI) antipsychotic treatment.
To describe patients attitudes and preferences towards oral versus LAI antipsychotic formulations and explore factors associated with their drug attitudes.
Two psychiatric hospitals in KwaZulu-Natal, South Africa.
A cross–sectional survey of 140 adult outpatients with schizophrenia spectrum disorders receiving LAI with or without oral antipsychotics (a total of 70) were compared to patients receiving oral antipsychotics only (
Of the 140 participants, 98 (70%) preferred the medication formulation currently prescribed, and 132 (94.3%) reported a positive drug attitude towards their antipsychotic medication. The adjusted regression analysis indicated that study participants who were currently on a formulation that matched their preference scored better on the DAI-30 than individuals with a mismatch in use and preference (
The majority of participants preferred their current oral and LAI formulation. Drug attitude was influenced by several factors, including matched medication use. Focused psychoeducation should be considered for newly diagnosed, lower socio-economic groups and patients with non-affective psychosis to improve drug attitude.
Schizophrenia is a chronic and disabling illness, with most patients experiencing multiple relapses during their illness; this risk is diminished by maintaining antipsychotic drug treatment.
The National Institute for Health and Care Excellence (NICE) guideline for medication adherence recommends that patients should be as involved as possible in decisions about the choice of medicines that are prescribed for them, and that clinicians should be aware that illness beliefs and beliefs about medicines influence adherence. The NICE guideline for schizophrenia further emphasises the importance of patient choice rather than specifically recommending a class or individual antipsychotic as first-line treatment.
Drug attitude is defined as a settled way of thinking or feeling about something;
Preference is defined as a greater liking for one alternative over another or other options.
Perceived well-being, spiritual causation, financial challenges and stigma were also associated with poor adherence to antipsychotic medication.
Studies from Africa on patient attitude to LAI compared to oral antipsychotic preparations are limited. To the authors’ best knowledge there is no study of patient perspective of LAI antipsychotics compared to oral preparation in Africa. The only study on attitude towards LAI antipsychotics in Nigeria was based on clinician attitude towards depot antipsychotics.
The aim of this study was to describe the attitude and preferences towards oral compared to LAI antipsychotic formulations among patients with a psychotic disorder (schizophrenia spectrum disorders) attending at either one of the two psychiatric hospitals in KwaZulu-Natal (KZN), South Africa.
The study objectives were to describe and compare the sociodemographic and clinical profiles of two groups of patients, those taking oral antipsychotic only and those taking LAI and/or oral antipsychotic medication, in relation to their drug attitudes and personal preferences to medication. In addition, associations between sociodemographic and clinical factors and total drug attitude score in patients receiving oral compared to LAI with or without oral antipsychotic preparation are described. The LAI group was subdivided into LAI only and those on combination of LAI and oral antipsychotic in sub-analysis.
A descriptive, cross-sectional study involving patient interviews, chart reviews and questionnaire interviews with adult outpatients receiving either LAI with or without oral antipsychotic, or oral antipsychotic medication only was conducted at two psychiatric hospitals in KZN. The data was collected over a seven-month period, from July 2018 to January 2019. A random sample of patients attending the hospital clinics and meeting inclusion criteria were invited to participate in the study.
The study was conducted at the two outpatient departments, at which 140 adult patients (aged 18 to 65 years) were randomly recruited from the outpatient services. Seventy patients were prescribed LAI with or without another oral antipsychotic medication, while 70 were on oral antipsychotic medication only. All participants were on antipsychotics for at least six months, had a diagnosis of a primary psychotic disorder based on the DSM 5 criteria for schizophrenia and schizophrenia spectrum disorders, did not require admission and were able to provide informed consent. The diagnosis was confirmed by clinical interview using the DSM 5 criteria and chart review. Participants were literate, able to speak English and provided written consent.
The two study sites are specialist psychiatric hospital services with inpatient and outpatient services that serve as the referral centres to the general hospitals in the eThekwini (site 1) and uMsunduzi (site 2) municipalities in KZN.
Of the 140 patients, 93 were from site 1 and 47 were from site 2. The first psychiatric hospital has approximately 900 adult outpatients per month and the second hospital has 450 adult outpatients per month.
The self-designed questionnaire was administered for this study, with data on demographic factors, such as age, gender, educational level, race, marital status and household monthly income being collected. The questionnaire included clinical factors such as primary psychiatric diagnosis (stratified into schizophrenia, schizoaffective disorder and other psychotic disorders), comorbid psychiatric diagnoses (as per DSM 5 criteria and chart review), comorbid medical disorders, number of admissions in past and treatment details, such as medication type and duration based on a review of the literature.
In addition, self-reported personal treatment preference towards oral or LAI formulation was also collated using a structured question with the option of selecting oral only or LAI only as their preferred treatment option. Participants selected their main reason for formulation preference. All interviews and chart reviews to confirm the clinical factors were conducted by the principal investigator. Interviews were conducted in English.
The DAI is a 30-item self–report inventory that focuses on the subjective effects of neuroleptic medication in patients with schizophrenia to understand the factors that influence drug attitude.
The DAI has been used in many developing countries, such as Ethiopia and Nigeria, as well as in studies outside Africa, including in the Clinical Antipsychotic Trial of Intervention Effectiveness (CATIE).
The DAI has 15 items that are scored as TRUE and 15 as FALSE in the case of a fully compliant patient (positive subjective response) as per the questionnaire. A correct answer to these items is scored as +1, while an incorrect response is scored as –1. The final score is the sum of the total of pluses and minuses. A positive total score suggests a positive subjective response or attitude, while a negative score means a negative subjective response or attitude. The maximum score is 30 if all responses are correct. The DAI score has been reported to predict adherence in schizophrenia,
The sociodemographic questionnaire and the DAI were pilot tested on 10 patients meeting the same inclusion criteria from the same study sites to determine its ease of understanding and the wording of questions. No problems were reported by patients with understanding the questions and the questionnaire was not modified.
Three analyses were conducted for this investigation to address the study objectives. Firstly, descriptive statistics were used to summarise the participants’ sociodemographic and clinical profiles (including attitude towards medication). Secondly, the sociodemographic and clinical profiles were compared between three groups of participants who were prescribed: (1) oral medication only, (2) LAI only or (3) both, based on chi-square (χ2) test.
Thirdly, multivariate regression models were used to investigate the association between concomitance of medication (i.e. matching of preferences and actual treatment) and the DAI scores using a regression model.
The model was adjusted by sociodemographic (gender, race, age, marital status, education and household income) and clinical covariates (psychiatric diagnosis, history of rehospitalisation due to medication discontinuation, number of past psychiatric admissions and duration of illness). Given the skewness of the DAI score, cubic data transformation was applied to normalise the distribution, which performed better than other transformation (e.g. square, square root, log). A significance of
The study was approved by the University of KwaZulu-Natal’s Biomedical Research Ethics Committee (Ref. No.: BE054/18). Permission was also obtained from the hospitals and the Department of Health.
Of the 140 participants, the majority were male (
Sociodemographic and clinical profiles of participants.
Variable | % | |
---|---|---|
Male | 97 | 69.29 |
Female | 43 | 30.71 |
Black people | 58 | 41.43 |
White people | 9 | 6.43 |
Indian people | 51 | 36.43 |
Mixed race people | 22 | 15.71 |
19–24 | 24 | 17.14 |
25–34 | 38 | 27.14 |
35–49 | 44 | 31.43 |
50+ | 34 | 24.29 |
Never married | 97 | 69.29 |
Married | 25 | 17.86 |
Divorced or widowed | 18 | 12.86 |
Grade 11 or lower | 67 | 47.86 |
Grade 12 | 31 | 22.14 |
Higher than 12 | 42 | 30.00 |
R2500 or lower | 39 | 27.86 |
R2501 – R6000 | 65 | 46.43 |
R6000 or higher | 36 | 25.71 |
Schizophrenia | 96 | 68.57 |
Schizoaffective disorder | 35 | 25.00 |
Other | 9 | 6.43 |
Yes | 73 | 52.14 |
No | 67 | 47.86 |
≤ 2 | 74 | 52.86 |
3–5 | 44 | 31.43 |
≥ 6 | 22 | 15.71 |
LAI only | 18 | 12.86 |
Oral only | 70 | 50.00 |
LAI and oral | 52 | 37.14 |
Oral | 96 | 68.57 |
LAI | 44 | 31.43 |
LAI, long-acting injectable antipsychotic treatment.
Of the 140 patients, 70 were on oral medication only, while in the LAI group (
Sociodemographic and clinical profiles of participants stratified by type of antipsychotic medication formulation.
Variable | LAI only ( |
Oral only ( |
LAI and Oral ( |
Total ( |
Test statistics | |||
---|---|---|---|---|---|---|---|---|
% | % | % | ||||||
χ2(2) = 0.56, |
||||||||
Male | 12 | 12.4 | 47 | 48.5 | 38 | 39.2 | 97 | |
Female | 6 | 14.0 | 23 | 53.5 | 14 | 32.6 | 43 | |
χ2(6) = 4.78, |
||||||||
Black people | 7 | 12.1 | 33 | 56.9 | 18 | 31 | 58 | |
White people | 1 | 11.1 | 4 | 44.4 | 4 | 44.4 | 9 | |
Indian people | 5 | 9.8 | 23 | 45.1 | 23 | 45.1 | 51 | |
Mixed race people | 5 | 22.7 | 10 | 45.5 | 7 | 31.8 | 22 | |
Total | 18 | 12.9 | 70 | 50 | 52 | 37.1 | 140 | |
χ2(6) = 9.58, |
||||||||
15–24 | 3 | 12.5 | 11 | 45.8 | 10 | 41.7 | 24 | |
25–34 | 4 | 10.5 | 20 | 52.6 | 14 | 36.8 | 38 | |
35–49 | 3 | 6.8 | 28 | 63.6 | 13 | 29.5 | 44 | |
50+ | 8 | 23.5 | 11 | 32.4 | 15 | 44.1 | 34 | |
χ2(4) = 2.67, |
||||||||
Single | 10 | 10.3 | 48 | 49.5 | 39 | 40.2 | 97 | |
Married | 5 | 20 | 12 | 48 | 8 | 32 | 25 | |
Divorced or widowed | 3 | 16.7 | 10 | 55.6 | 5 | 27.8 | 18 | |
χ2(4) = 2.42, |
||||||||
Grade 11 or lower | 10 | 14.9 | 34 | 50.7 | 23 | 34.3 | 67 | |
12 | 3 | 9.7 | 18 | 58.1 | 10 | 32.3 | 31 | |
Higher than 12 | 5 | 11.9 | 18 | 42.9 | 19 | 45.2 | 42 | |
χ2(4) = 3.18, |
||||||||
R2500 or lower | 7 | 17.9 | 15 | 38.5 | 17 | 43.6 | 39 | |
R2501 – R6000 | 7 | 10.8 | 35 | 53.8 | 23 | 35.4 | 65 | |
R6001 or higher | 4 | 11.1 | 20 | 55.6 | 12 | 33.3 | 36 | |
χ2(4) = 2.33, |
||||||||
Schizophrenia | 12 | 12.5 | 48 | 50 | 36 | 37.5 | 96 | |
Schizoaffective disorder | 6 | 17.1 | 16 | 45.7 | 13 | 37.1 | 35 | |
Other | 0 | 0 | 6 | 66.7 | 3 | 33.3 | 9 | |
χ2(2) = 1.56, |
||||||||
Yes | 11 | 15.1 | 33 | 45.2 | 29 | 39.7 | 73 | |
No | 7 | 10.4 | 37 | 55.2 | 23 | 34.3 | 67 | |
χ2(4) = 1.05, |
||||||||
≤ 2 | 9 | 12.2 | 40 | 54.1 | 25 | 33.8 | 74 | |
3–5 | 6 | 13.6 | 20 | 45.5 | 18 | 40.9 | 44 | |
≥ 6 | 3 | 13.6 | 10 | 45.5 | 9 | 40.9 | 22 | |
χ2(2) = 27.04, |
||||||||
Oral | 7 | 7.3 | 62 | 64.6 | 27 | 28.1 | 96 | |
LAI | 11 | 25.0 | 8 | 18.2 | 25 | 56.8 | 44 |
LAI, long-acting injectable antipsychotic treatment.
When all groups are combined, 98 (70) of the 140 participants preferred the medication formulation currently prescribed. Next, we investigated the reasons of treatment preference. Of the 44 participants who preferred LAI only (not shown in
The mean DAI score was 16.65 (SD 9.53) with a median score of 18 and IQR of 12 for all participants. The majority (
The total number of correct responses to the individual items on the DAI (
The most common incorrectly answered questions for the LAI group were question 12 (‘Medication makes me feel tired and sluggish’ – correct response false as per DAI) and question 27 (‘I am given medication to control behaviour that other people [not myself] don’t like’). Question 27 of the DAI was also most commonly incorrectly answered by the participants in the oral antipsychotic only group. Question 20 (‘It is unnatural for my mind and body to be controlled by medication’) of the DAI was incorrectly answered by 52 participants in total, 28 (40.0%) on LAI antipsychotic and 24 (24.3%) participants on the oral antipsychotic.
The analysis for the differences in attitude on the individual DAI statements (the selected 10 that are included in the DAI-10) between the participants in the two groups (oral compared to LAI with or without oral) showed no statistically significant differences between them (analysis not shown).
The association between the sociodemographic and clinical variables and the total DAI score using a regression model are summarised in
Association between medication preference and Drug Attitude Inventory scale score using regression model.
Variable | Adjusted β | Standard error | ||
---|---|---|---|---|
[Male] | - | - | - | - |
Female | −7.90 | 14.27 | −0.55 | 0.58 |
[Black people] | - | - | - | - |
White people | 24.89 | 27.76 | 0.90 | 0.37 |
Indian people | −7.91 | 17.01 | −0.47 | 0.64 |
Mixed race people | −30.50 | 19.16 | −1.59 | 0.11 |
[19–24] | - | - | - | - |
25–34 | −7.12 | 19.22 | −0.37 | 0.71 |
35–49 | 10.08 | 22.71 | 0.44 | 0.66 |
50+ | −1.35 | 28.08 | −0.05 | 0.96 |
[Never married] | - | - | - | - |
Married | −17.79 | 17.90 | −0.99 | 0.32 |
Divorced or widowed | − |
20.41 | −3.13 | < 0.01 |
[Grade 11 or lower] | - | - | - | - |
12 | −6.22 | 15.44 | −0.40 | 0.69 |
Higher than 12 | 1.27 | 15.35 | 0.08 | 0.93 |
[R6000 or higher] | - | - | - | - |
R2500 or lower | 7.93 | 17.74 | 0.45 | 0.66 |
R2501 − R6000 | 37.85 | 16.48 | 2.3 | 0.02 |
[Schizoaffective disorder] | - | - | - | - |
Schizophrenia | − |
15.09 | −2.55 | 0.01 |
Other | −6.83 | 28.36 | −0.24 | 0.81 |
[Yes] | - | - | - | - |
No | −3.39 | 14.09 | −0.24 | 0.81 |
[≤ 2] | - | - | - | - |
3–5 | −3.37 | 15.84 | −0.21 | 0.83 |
≥ 6 | −22.43 | 20.14 | −1.11 | 0.27 |
In years | 1.93 | 0.92 | 2.11 | 0.04 |
[Divergent] | - | - | - | - |
LAI individual preferring LAI | 49.61 | 23.28 | 2.13 | 0.04 |
Oral individual preferring oral | 27.35 | 13.53 | 2.02 | 0.04 |
LAI, long-acting injectable antipsychotic treatment.
Note that adjusted β are transformed. Caution is warranted when interpreting.
The study aimed to describe the drug attitude and preferences towards oral compared to LAI antipsychotics among patients with a psychotic disorder (schizophrenia spectrum disorders as per DSM 5) attending either one of the two psychiatric hospitals in KZN. The DAI was utilised to measure drug attitude and may indirectly predict patient adherence to psychotropic medication.
Long-acting injectable formulation is usually viewed as a monotherapy modality to improve medication adherence adherence; however, some evidence still exists that combination with oral antipsychotic formulations does occur, potentially representing a form of polypharmacy.
Patients’ preferences towards antipsychotic medication formulations in general – whether LAIs or oral formulations – are associated with several factors, including current formulation, symptom remission, functioning, side effects and treatment outcomes.
In this study, most participants reported a positive drug attitude score, which is consistent with the literature.
This current study showed an association between negative drug attitude and divorced or separated status, suggesting that individuals with a negative subjective response or attitude are more at risk of non-adherence.
While there were no studies assessing marital status and drug attitude, studies assessing the association of marital status and medication adherence (predicted by the DAI)
The literature reports that financial burden and unemployment were associated with medication non-adherence in patients with schizophrenia,
More positive drug attitudes have been consistently found to be associated with the ‘insight’ into the presence of a mental disorder.
This finding highlights the need to intensify psychoeducation efforts for those diagnosed with schizophrenia to potentially improve adherence.
The association between duration of illness and drug attitude score towards psychotropic medication has been established, and shows that the longer the duration of an illness the more favourable the subjective attitude to medication.
The current study failed to find an association between drug attitude score and history of rehospitalisation due to treatment discontinuation or number of admissions. This is not consistent with the literature,
Several limitations have been identified in this study, including that it was urban and hospital-based, which may limit the findings generalisations, although both psychiatric hospitals receive patients from a large referral area that includes rural residents. There may be also patient bias in a hospital-based samples as these patients are often more difficult to treat, hence are prone to relapse, compared to patients referred back to local clinics for further care. The lack of validation of the DAI tool in the local setting and recruiting only English-speaking participants in the study limits the findings further. The sample size may additionally limit the assessment for associations. The cross-sectional nature of the study limits the findings, which may be better assessed longitudinally, as patients surveyed were adherent and attending clinic at the time of survey. There may be under-reporting of non-adherence of medication, as the study relied on subjective reports. In our setting, patients on LAI antipsychotic only were limited in number, as many of those who were assessed as using LAI, were also on oral antipsychotic medication; however, this possibly provides a more realistic picture of current local practice.
Furthermore, this study is limited as insight and side effects were not measured and hence, their effect on DAI was not assessed. Finally, it will be useful to have included collateral information from the caregiver to get more objective data on patients’ drug attitude.
The key findings of the study are that most participants reported a positive drug attitude towards their current antipsychotic formulations. The majority of participants preferred oral compared to LAI formulation. There was a positive association between drug attitude score and subjective formulation preference, and participants who were on a formulation that matched their preference scored better on the DAI compared to individuals with mismatch in use. The findings suggest the need for considering drug formulation preference and more focused psychoeducation in groups vulnerable to poor drug attitude, such as the newly diagnosed, unemployed, single, those with schizophrenia and those on regimens with polypharmacy.
Further research is required to better assess patient drug attitude in larger longitudinal studies, and to consider interventions to improve drug attitude and possibly adherence.
The authors thank Prof. B. Chiliza, Dr S. Ramlall and Dr J. Brooker for facilitating approval of the research. The authors also appreciate the help of various nursing staff who facilitated the process of collecting information.
The authors have declared that no competing interests exist.
K.R.R. conceptualised the study, collected the data and wrote the final draft. A.T. analysed the data. S.P. supervised the study and edited the article.
Biomedical Research Ethics Committee, University of KwaZulu-Natal, REC-290408-009.
This research received a CHS (College of Health Sciences) scholarship grant from the University of KwaZulu-Natal. A.T. was supported by the South African Medical Research Council (MRC-RFAUFSP-01-2013/UKZN HIVEPI) with funds from South African National Treasury under its Economic Competitiveness and Support Package.
The data that support the findings of this study are available on request from the corresponding author, K.R.R. The data are not publicly available due to restrictions (e.g. they contain information that could compromise the privacy of research participants).
The content is solely the responsibility of the authors and does not necessarily represent the official views of the South African Medical Research Council or South African National Treasury.