Attitudes toward psychotropic medications among patients with chronic psychiatric disorders and their family caregivers

Participants

The sample for the study comprised of 200 patients and their family caregivers with chronic psychiatric disorders who are attending the psychiatry outpatient services.

Patients were included if they were more than 18 years of age, were suffering from any chronic psychiatric disorder of more than a year's duration, were receiving psychotropic medications for at least 1 year, were able to read Hindi or English, and consented to participate in the study. All patients were “clinically stable”, i.e. there had been no major change in their clinical condition in the 3 months prior to inclusion, and not more than a 50% hike or reduction of dosages of psychotropics during this period. Patients who were acutely ill, uncooperative, aggressive or actively suicidal, or cognitively impaired were excluded.

Family caregivers of these patients were included if they were more than 18 years of age, were living with and involved in patient's care for the past 1 year or more prior to inclusion, were able to read Hindi or English, and consented to participate in the study. Family caregivers who were themselves suffering from any mental illness or mental retardation were excluded.

Assessments

Apart from recording demographic, clinical and treatment data, the principal task of this study was to assess attitudes toward psychotropic medications among patients and their family caregivers. There are several well-validated scales for assessment of attitudes among patients including the Drug Attitude Inventory (DAI),[18] the Rating of Medication Influences scale (ROMI).[19] the Attitudes toward Neuroleptic Treatment (ANT)[20] questionnaire, and the Antidepressant Compliance Questionnaire (ADCQ).[21] However, for several reasons these scales did not suit the purpose of this study. Firstly, all these scales were meant for assessing attitudes among patients, and not meant to be used for assessing attitudes of caregivers. Studies assessing caregivers’ attitudes have devised their own questionnaires, but these have not been properly validated.[1416] Secondly the DAI, the ROMI and the ANT questionnaire are meant to assess attitudes toward antipsychotics. They have thus been mainly used among patients with psychosis or schizophrenia, and only occasionally among patients with depressive or bipolar disorders.[2223] Finally, in our experience some of the items of these scales such as the DAI are not well understood by Indian patients,[242526] while other common culturally derived beliefs and attitudes are, understandably, not incorporated in these scales.

Accordingly, a questionnaire, named Self Report Attitude Toward Psychotropic Medications Questionnaire, was devised to assess attitudes toward psychotropic medications among patients with a wide range of chronic psychiatric disorders including psychotic disorders, mood disorders (both major depressive and bipolar disorders), and other chronic psychiatric disorders. Moreover, the questionnaire was devised in such a way that it could also be used to elicit attitudes of family caregivers toward psychotropic medications. For this purpose, a 7-item questionnaire for measuring attitudes toward psychotropic medications or antipsychotics among laypersons, as well as professional and non-professional caregivers was adapted for use in this study.[2728] Firstly, the wording of the items was modified to elicit attitudes toward psychotropic medications in general and not just antipsychotics. Secondly, the scoring pattern of the scale was simplified. In the original version of this instrument, each statement is rated on a five-point scale from: 1- ‘completely agree’ to 5- ‘completely disagree’.[28] While the original Likert format of the scale was maintained, items were scored on a three-point scale with three options for the items eliciting positive attitudes including: 1 - “strongly disagree”. 2- “agree to an extent”, and 3- “strongly agree”. For the items assessing negative attitudes, the pattern of scoring was reversed, which ensured that a higher score always reflected more positive attitudes on each item (1-3) and for the total score of the scale (range 18-54). Finally, additional items which reflected common culturally derived attitudes toward medications were incorporated after discussions with other mental health professionals, patients and their caregivers. The outcome of this process resulted in an 18-item questionnaire in English, which was translated to Hindi (the local language) using standardized methodology proposed by the World Health Organization.[29] All efforts were made to keep the questionnaire as short and simple as possible to enhance its ease of use. The Hindi version of the questionnaire was again reviewed by a panel of experts and was administered to the patients and their caregivers to assess the language and wording of the questions. All the suggestions made during this phase were incorporated by arriving at a consensus. Of the 18 items of the questionnaire, 10 items assessed negative attitudes and 8 items assessed the positive attitudes toward psychotropic medications. The questionnaire was designed to be used as a self-report measure. Before asking the participants to complete the questionnaire, clear instructions were given about the purpose of the questionnaire and the response categories. Once these instructions were understood, most participants had no difficulty in completing the questionnaire and took about 4-6 minutes to complete the questionnaire.

Data analysis

Data were analyzed using the Statistical Package for Social Scientists, version fourteen (SPSS-14).[30] Frequency, percentage, mean, and standard deviation were calculated for the descriptive data. Comparisons were done using the Chi-square and Fishers’ exact test for the nominal and ordinal variables, and t-tests for the continuous variables. Exploratory factor analysis was carried out using a principal components analysis. Principal components analysis was chosen as the primary aim was to determine the minimum number of factors that would account for the maximum variance in the data. Initially a correlation matrix was prepared. The Kaiser Meyer Olkin measure of sampling adequacy and the Bartlett's Test of Sphericity were assessed. Once it was confirmed that the data lent itself to factor analysis, the number of factors derived on the basis of un-rotated matrix were evaluated. Next the data were subjected to orthogonal rotation using the varimax technique to increase interpretability and interpretation of the results. Loadings of ≥0.4 of an item on a particular factor were considered to be significant loadings, and the item was included in that factor. When the variable loaded at 0.40 or above on more than one factor, it was assigned to the factor on which it had the higher loading. For determining the optimum number of the factors, the Kaiser-Guttman Rule[31] was used. Only factors with Eigen value of more than 1 and loadings that were ≥0.4 were retained to render the extracted factors meaningful and interpretable. A scree plot was also generated, and previous literature was referred to determine the factor solutions that would provide best fit for the data.

Attitude toward psychotropic medications patients and family caregivers

The responses on different attitude items, the total scores and scores on positive and negative attitude items among patients and family caregivers are depicted in Table 1.

Table 1 Attitude of patients and caregivers toward psychotropics

As shown in Table 1, on the 8 items assessing positive attitude toward psychotropics, responses of 58-78.5% of patients and 54.5-83.5% of caregivers reflected favorable attitude toward psychotropics. With regards to the 10 items assessing the negative attitude toward psychotropics, responses of 8.5-79.5% of patients and 5-82% of caregivers reflected that they had favorable attitude toward psychotropics.

The mean total score of patients was 43.52 ± 5.02 (range 23-53) and that of caregivers was 44.01 ± 4.97 (range 30-53), out of the maximum obtainable score of 54.

Mean weighted positive attitudes scores (obtained by dividing the total raw score by total number of items assessing the positive attitude, i.e. 8) for patients was 2.64 ± 0.34 (range 1.38-3), while mean weighted negative attitudes scores (obtained by dividing the total raw score by total number of items assessing the negative attitude, i.e. 10) was 2.23 ± 0.33 (range 1-3). This indicated that positive and negative attitudes toward psychotropic medications were equally prevalent among patients. Similarly, for caregivers mean weighted positive attitudes scores was 2.67 ± 0.32 (range 1.62-3), while mean weighted negative attitudes scores was 2.25 ± 0.353 (range 1-2.9), which also indicated that caregivers had an equal share of both positive and negative attitudes toward psychotropic medications.

Comparisons of attitudes between patients and caregivers revealed differences only on one item, i.e. psychotropics “Don’t cure, but lead to substantial improvement”. Thus, attitudes toward medications were similar among patients and their family caregivers.

Factor analysis of attitudes toward psychotropic medications among patients and family caregivers

Separate factor analyses were run for the data generated from the patients and the caregivers. Additionally, the final factor analysis included data of both patients and the caregivers. The details of the various factor analyses are given in [Tables 2 and 3]. In all the factor analyses, the initial unrotated factor analysis yielded a five-factor structure, which explained 53.59% to 54.73% of variance. The scree plot for the factor analysis of the scale based on responses of patients is depicted in the accompanying figure.

Table 2 Factor analysis of the scale

Table 3 2-Factor structure of the scale

Factor solutions

Factor loadings, the total variance explained by the different numbers of factors, the Scree plot [Figure 1] and previous literature were all used to determine the factor solutions that provided best fit for the data. The two-factor solution (positive and negative attitude factors) was in keeping with many of the earlier studies on attitudes with the DAI.[63233] However, the variance explained by the two-factor solution was the least.

There was very little difference between the loadings on different factors in the three-, four-, and five factor solutions. The scree plots appeared to tail off at the four or five factors. Studies of the DAI have also yielded between three to seven factors,[183334] while a three to five factor structure has also been proposed for the ROMI scale,[19] and a three factor solution has been proposed for the ANT questionnaire.[20] Therefore, a four-factor solution was also considered as ideal fit for the data.

Close

Figure 1

Scree plot showing the factor analysis solution for the data on attitude toward psychotropics

Export to PPT

Details of factor loadings for the two-factor solution have been included in Table 3, and those for the four-factor solution have been depicted in Table 4.

Table 4 Four-factor structure of the scale

The two-factor solution

In the two-factor structure for patients, factor one comprised of items 1 to 8, except item 3, which though loaded on to factor one, just fell short of the cut-off of 0.40. In addition, item 11 demonstrated an inverse loading on this factor. Thus, factor one appeared to be a “positive attitudes” factor. However, there were two important discrepancies. Item 10 (“Psychotropic medications are unnatural and poisonous substances, which are harmful”), which reflected negative attitudes also loaded on to factor one. Item 15 (“Psychotropic medications are not necessary for treatment of mental illness, because mental illnesses can be controlled by other means too”), which also reflected negative attitudes loaded on to both factors, though its loading on factor one was somewhat greater. It was not clear why these two items should load on to factor one, which essentially comprised of positive attitudes. It is possible that there was some ambiguity in the wording of these items, which led to these discrepant loadings. Factor two comprised of four items (items 9, 10, 16 and 17), which all reflected negative attitudes. Thus, factor two was the “negative attitudes” factor.

This positive (factor one) and negative (factor two) attitudes distinction was more apparent in the case of caregivers with all positive attitude items (except item 3 and 5), and all negative attitude items (except item 11) loading on to factor one and two respectively.

However, this positive-negative attitudes distinction was best illustrated by the results of the factor analysis of the combined sample of 200 patients and 200 caregivers. In this analysis, factor one comprised of all the positive attitude items (from item 1 to item 8) and the inverse of a negative attitude item (item 11), while factor two comprised of all negative attitude items, except item 11.

The four-factor solution

In the four-factor solution for patients, factor one comprised of all items reflecting positive attitudes from item 1 to item 8, except item 3, which loaded separately on to factor four. Analysis of individual items in factors one and four appeared to indicate that a risk-benefit approach was being followed by patients, in which the benefits of medication was contrasted with the risks to arrive at a positive attitude regarding medication. Factors two and three both reflected negative attitudes. Factor two was comprised of items that focused on harm caused by medications (items 10, 12, 17 and 18), and one item reflecting the lack of necessity for these medications (item 15 - “Psychotropic medications are not necessary for treatment of mental illness, because mental illnesses can be controlled by other means too”). Factor three was exclusively comprised of items that reflected the harm caused by these medications. Moreover, the items comprising this factor also possibly reflected common culturally derived attitudes toward allopathic medications, since the items focused on the risk of dependency (item 9), medications making the body unnaturally hot or cold (tem 13), or draining energy and causing weakness (item 16), and on medications being very expensive (item 14).

Among caregivers these four factors were again replicated with some minor differences. For example item 10, which reflected negative attitudes loaded on to factor one, and factor 5, which reflected positive attitudes loaded both on to factor three and factor four. Apart from these discrepancies the factor structure was essentially similar to that obtained among patients with factors one and four as positive attitude factors, derived from a risk-benefit approach that favored medications, and factors two and three being negative attitude factors, which focused on concerns regarding medications such as potential harm, expense and lack of necessity.

The results also indicated that the four-factor structure was the best fit for the data obtained from the combined sample of 200 patients and 200 caregivers.