Treatment compliance and noncompliance in psychoses

Abstract

Background:

Compliance or noncompliance with treatment significantly influences course and outcome of psychiatric disorders. While noncompliance has been extensively researched, compliance has received less attention. The current study was conducted to elicit reasons for compliance and noncompliance in patients having psychoses attending psychiatric clinics.

Materials and Methods:

A total of 196 compliant and 150 noncompliant patients were interviewed using self-designed tools to elicit sociodemographic data, details of illness, and treatment. Factors contributing to compliance and noncompliance were grouped under illness-related, clinician-related, medication-related, family-related, and economic-related domains and compared.

Results:

Compliance was significantly more in females and middle- and high-socioeconomic status patients. They had less substance use, high physical comorbidity, high attendance in the outpatient department, and better remission. Clinician-related, family-related, and medication-related domains were contributing more to compliance whereas illness-related and economic-related domains seemed to have more bearing on noncompliance.

Conclusions:

Compliance and noncompliance are determined multidimensionally. Domains related to clinician, family, and medications have to be reinforced to enhance compliance. Illness-related and economic domains have to be resolved to reduce noncompliance.

INTRODUCTION

There have been dramatic improvements in pharmacotherapy of psychotic disorders in the past 60 years contributing to reduction of symptoms, prevention of relapse, and improvement in social functioning.[1] Despite these gains, treatment of psychotic disorders remains a major challenge. The actual effectiveness of antipsychotic medications is well below the efficacy obtained from clinical trials when relapse rates are considered.[2] One of the reasons for this state could be noncompliance to treatment. Compliance is defined as the extent to which a person's behavior coincides with medical or health advice.[3] Noncompliance can be defined as opposite of compliance. There has been an attempt to divide noncompliance into primary (not buying or receiving the medicines) and secondary (not complying with the instructions regarding dosage, frequency, and duration of medication intake).[4] In recent times, the terms compliance and noncompliance have been considered as clinician oriented. To shift the emphasis onto patients, the terms adherence and nonadherence have been suggested. However, in clinical practice, these terms are used interchangeably.

Noncompliance rates in schizophrenia vary widely, ranging from 20% to 89%.[5] Various reasons have been cited in literature for noncompliance. These include poor insight, side effects of medicines, poor remission of symptoms, and poor therapeutic alliance.[6,7,8,9] It also includes stigma associated with the illness,[9,10] poor family support,[11] ignorance about need to continue treatment,[12] and economic reasons.[9,13] In addition, when patients improve, they may not feel the need to continue medications anymore.[12] In Indian studies, noncompliance has been found to be related to lack of knowledge, financial difficulties, side effects, and no improvement.[14,15] Distance to hospital, lack of caregivers, poor insight, and lack of time have also been cited as reasons for noncompliance.[15] It is also reported that there is high prevalence of substance abuse in schizophrenia in noncompliant patients.[16] Other contributing factors are low priority accorded to health by patients and their caregivers, higher importance to economic activities, tendency to deal with problems only when they become acute and high emphasis on doctors’ exclusive role in alleviating symptoms of the patient.[17] Noncompliance is strongly related to clinical outcomes such as relapse, rehospitalization, and suicide attempts in schizophrenia[18] and bipolar disorder.[19] It also results in poor quality of life and financial burden with about 40% of total costs of illness being attributed to rehospitalizations.[20]

The reasons cited for compliance are wish to lead a normal life, fear of illness, advice of family and friends, and clinical improvement with treatment.[12] Attitude of family and friends has been found to be related to both compliance and noncompliance.[9,12]

Whereas noncompliance has been fairly well researched, compliance has received less attention. There have been very few studies which have looked at the reasons for compliance and noncompliance simultaneously.[12] The current study assesses and compares reasons for compliance and noncompliance simultaneously in patients having psychoses.

MATERIALS AND METHODS

Patients of both genders with a minimum age of 18 years having a primary diagnosis of psychoses attending psychiatric outpatient department (OPD) formed the sample of the study. The patients were divided into two groups – compliant and noncompliant based on their visits to treatment centers in the preceding 1 year. Compliant patients were defined as those who took medications for at least 80% of the days in the past 1 year.[21] The rest were considered as noncompliant. This was determined from the follow-up records. Patients aged below 18 years, those with diagnoses other than psychotic disorders and whose medical records were inaccessible or incomplete were excluded from the study.

After obtaining informed consent, a minimum of 100 consecutive patients, at least 50 each of compliant and noncompliant were included in the study from three psychiatric centers. The three centers were government general hospital attached to a medical college, private medical college hospital, and private psychiatric hospital. The patients were drawn from three different setups, so that reasons for compliance and noncompliance could be generalized across different setups to some extent. The sample size was 112 patients from government general hospital (62 compliant and 50 noncompliant), 100 from private medical college hospital (50 compliant and 50 noncompliant), and 134 patients from private psychiatric hospital (84 compliant and 50 noncompliant). Thus, a total of 196 compliant and 150 noncompliant patients were enrolled into the study. The study was conducted from October 2015 to March 2016.

Sociodemographic data, details of illness, and treatment history were noted for each participant in a semistructured pro forma specifically designed for the study. A 26-item questionnaire to elicit reasons for compliance & noncompliance was devised based on reviewed literature and authors’ clinical experience. An allowance was made for additional reasons given by the patients. The items were grouped into illness-related, clinician-related, medication-related, family-related, and economic-related domains and numbered as 1, 2, 3, 4, and 5, respectively. The questionnaire was administered to patients individually. The statements had to be answered as either “Yes” or “No.” “Yes” responses were scored as “1” and “No” responses were scored as “0.”

Sociodemographic variables, illness variables, treatment variables, and factors contributing to compliance and noncompliance were analyzed and compared between compliant and noncompliant groups.

Statistical analysis was done using SPSS version 22 (IBM). Mean and standard deviation were calculated for continuous variables and proportions for categorical variables. Comparisons of mean values between two groups were analyzed using student's t-test unpaired. Comparison of mean value within the group was analyzed using student's t-test paired. Mann–Whitney test was used where the data failed normality test. To compare the association between groups, Chi-square test was used. P < 0.05 was considered statistically significant.

RESULTS

The sample consisted of 346 patients; 196 compliant and 150 noncompliant. Table 1 summarizes sociodemographic variables. Compliance was observed to be significantly more in females (58.67%) and those from middle- and high-socioeconomic strata (37.24%).

Table 1.

Sociodemographic variables in compliant and noncompliant groups

Table 2 shows the psychiatric diagnoses, comorbidities, substance use, family history of psychoses, and current clinical status. The most common diagnosis in both the groups was bipolar affective disorder (mania and depression), followed by schizophrenia spectrum disorders (schizophrenia, delusional disorder, schizoaffective disorder) and depressive disorders (recurrent depressive disorder, major depressive disorder with psychotic symptoms). Other diagnoses included organic psychosis, postictal psychosis, and alcohol-induced psychotic disorder. Substance use was found to be significantly more in noncompliant group (54%), whereas physical comorbidities were found to be significantly more in compliant group (21.96%). The two groups differed significantly regarding their current clinical status. Most of compliant patients were in complete remission (29.59%) or at least partial remission (66.84%), but majority of noncompliant patients were symptomatic (36.67%) or only partially remitted (60%).

Table 2.

Illness characteristics in compliant and noncompliant groups

Adherence to follow-up appointments was significantly more in compliant group than noncompliant group (P < 0.000). Follow-up visits by patients’ attendants to procure medicines on behalf of the patients (treatment by proxy) were significantly higher in compliant group (P < 0.000) [Table 3].

Table 3.

Comparison of compliant and noncompliant groups

Table 4 shows the domain-wise analysis of items associated with compliance and noncompliance. Mean scores of each domain were compared within and between groups. Total score and scores on domains 1, 3, 4, 5 were significantly higher in compliant group as compared to noncompliant group. Scores on domain-2 did not differ between groups.

Table 4.

Comparison between compliant and noncompliant groups

As different domains had different number of items, the mean scores of each domain were converted into percentages, so that they become comparable. Ranks were assigned to these percentages. In the compliant group, higher ranks indicated higher compliance. The domains 2 and 4 were related to high compliance (rank 1 and 2), domain 3 was related to moderate compliance (rank 3), and domains 1 and 5 were related to low compliance (ranks 5 and 4).

In the noncompliant group, higher ranks suggested higher noncompliance. The domains 1 and 5 were related to high noncompliance (rank 1 and 2), domain 3 was related to moderate noncompliance (rank 3), and domains 4 and 2 were related to low noncompliance (ranks 4 and 5). The domains which contributed more to compliance contributed less to noncompliance.

Item-wise comparisons in domains were done within and between groups [Table 5]. All the individual items in domain-1 were significantly positive in compliant group than in noncompliant group. Items of good doctor-patient relationship in domain-2, side effects in domain-3, family support in domain-4, and affordability of medications and time to attend OPD in domain-5 were significantly positive in compliant group than noncompliant group. However, the item that “no delay in receiving treatment at hospital” in domain-2 elicited significantly more positive responses in noncompliant group as compared to compliant group. Regarding the items which were not statistically different between the groups, the extent of their contribution to compliance or noncompliance was determined on the basis of the percentages of responses. Those items which were positively responded by more than 50% of the patients were considered to play equal role in compliance and noncompliance. Those items which were positively responded by < 50% of the patients were considered to play lesser role in compliance and noncompliance.

Table 5.

Item wise comparison between compliant and noncompliant groups

The item clinical improvement was found to contribute to both compliance and noncompliance for different reasons. 61.22% of compliant patients reported that they continued treatment as they felt that the treatment was effective while 52% of noncompliant patients reported that the clinical improvement made them think that medications are no longer required. If there was no improvement with initial treatment, it was reasoned by some that there was no need to continue ineffective treatment. This led to noncompliance without a chance to change the treatment regimen by the clinician.

DISCUSSION

Noncompliance is not limited to psychiatric patients. It occurs in all branches of medicine. The figures of noncompliance in nonpsychiatric patients in some Indian studies are HIV-30%,[22] tuberculosis-6 to 50%[23,24,25] multidrug-resistant tuberculosis-19%,[26] hypertension-28.9%,[27] diabetes mellitus-42.3%,[28] head and neck cancer-23.5%,[29] cervical cancer-61.1%,[30] curative cancer treatment-33.33%,[31] lung cancer-36%.[32] Even in studies conducted abroad noncompliance has been reported in various medical conditions; tuberculosis global noncompliance rate-61.7%,[33] HIV optimal treatment-68%,[34] hypertension-30.5%,[35] diabetes mellitus (new detection)-20.6%.[36] It has been reported that adequate adherence is found only in the 1^st year of treatment in breast cancer patients.[37] Thus, noncompliance is common in many chronic medical illnesses too.

Psychiatric disorders may have different perspective from that of chronic medical illnesses. Yet, psychoses are a group of chronic psychiatric disorders needing long-term medication. Hence, factors of noncompliance in them may have relevance to noncompliance in other chronic medical conditions.

Compliance and noncompliance mainly have three research questions: (1) extent of noncompliance, (2) determinants of noncompliance, (3) strategies to improve compliance and decrease noncompliance. This article deals with the second question.

Female patients in the present study were more compliant than males which is not in keeping with the findings of previous studies.[14] This might be so because females in this sample were mostly homemakers or agriculturists who had time to follow-up at hospital regularly and probably; they were also expected to resume their household responsibilities at the earliest. In the present study, compliance was found to be significantly more in patients from middle- and high-socioeconomic status. Financial status and awareness about health might have contributed to this finding. In line with this previous studies have observed higher noncompliance rates in lower socioeconomic strata.[38,39] In this study, there was no difference between the compliant and noncompliant group with respect to age, education, occupation, religion, family type, place of residence, and marital status. In contrast, other studies have observed that young age,[38] unmarried status,[40] lower educational qualification,[38,41] and joint family status[14] are associated with noncompliance.

The common psychiatric diagnosis in the present study was bipolar disorder followed by schizophrenia spectrum disorders. Similar finding has been noted in another Indian study.[14] Compliance did not differ significantly with respect to the diagnostic subgroups or a positive family history. The compliant group had significantly higher comorbidity of chronic physical illness. Probably, in them, hospital visits may serve a dual purpose of consultation for medical and psychiatric illnesses. Therefore, the treating clinician should give equal importance to treatment of patients’ physical as well as psychiatric conditions. As reported in earlier studies, in this study too, substance use was significantly associated with noncompliance.[16,42]

Neither total duration of illness nor duration of treatment was different between the two groups. As expected, compliant patients attended OPD as and when called whereas noncompliant patients did not. “Treatment by proxy” was significantly higher in compliant group. This is a flexible, low cost, and patient-friendly policy which could improve compliance.

Significantly more of the compliant patients had achieved complete or at least partial remission. This may mean that remission associated with compliance may boost confidence in treatment. A previous study has found that compliance is associated with lesser relapses and lesser rehospitalizations.[42] However, remission can also make some patients complacent and stop treatment. Some noncompliant patients remained symptomatic or only partially remitted. Association of lack of remission with noncompliance has been reported by Moritz et al.[9,12] Lack of remission can frustrate some patients and their attendants resulting in loss of faith in treatment and render them noncompliant. On the other hand, nonremission can goad them to continue treatment to achieve remission. Thus, clinical improvement determining compliance and noncompliance may be influenced by the perception of the patients and attendants. This stresses the role of clinicians in educating patients and their attendants about the need for continued treatment to prevent relapse and reasonable time required for remission of symptoms.

Inaccessibility to hospital services has been cited as a reason for noncompliance in previous studies.[15,43] In this study, there was no difference between the two groups with respect to distance to the hospital or transportation charges. Thus, these factors may not distinguish between compliance and noncompliance.

In domain-wise analysis, illness-related, medication-related, family-related, and economic-related domains were significantly associated with compliance than noncompliance. Clinician-related domain did not differ between groups. Identical scores in clinician-related domain between groups can imply that it is equally important in both the groups. Systematic reviews of research in this area have grouped the factors into different areas but have not statistically compared between groups.[44,45]

In the present study, clinician-related and family-related domains were contributing more to compliance (ranks 1 and 2) and less to noncompliance (ranks 4 and 5). Medication-related domain was related to moderate compliance and noncompliance (rank 3). Illness-related and economic domains seemed to have less bearing on compliance (ranks 4 and 5) than noncompliance (ranks 1 and 2).

In addition to domain-wise comparison, individual items were also compared between two groups.

In line with earlier research, there was positive influence of good doctor-patient relationship on compliance.[46,47] Communication skills of doctors, promptness of service, and personal attention of doctor may also have a significant role in compliance and noncompliance. These points have not been alluded in earlier studies.

The findings of present study concurred with earlier studies that good family support and positive attitude of family members were significantly associated with high compliance[12] and unfavorable attitude with noncompliance.[46] The influence of family on compliance is a modifiable factor which can be achieved through psychoeducation.[11,48,49] Implementation of this measure is beset with operational difficulties in countries with limited workforce. Stigma of mental illness affecting compliance and noncompliance has been reported in previous studies.[9,48,50] However, stigma of relapse was found to play a lesser role in compliance and noncompliance in our study.

In medication-related domain, previous studies have reported that side effects of medicines as barrier to compliance[12,14,15] and once daily drug dosage and long-acting injectables in place of oral medicines as facilitators of compliance.[9,48,50] Present study, while concurring with these factors found that factors of ease of availability of medicines, lesser number of medicines, ease of swallowing medications, and easy regimen of medications also played a role in compliance.

In illness-related domain, awareness about illness, need for long-term treatment, wish to lead a normal life, and fear of symptom deterioration were associated with compliance as has been reported in earlier studies.[12,15,46,51,52] Thus, imparting insight about the nature of illness, emphasis about the need for prolonged treatment is crucial to reduce rates of noncompliance. Disparity between explanatory models of illness held by doctors and patients is also to be minimized to improve compliance.[4]

In economic domain, impact of the item “availability of medicines for free” in compliance and noncompliance cannot be commented because only one of the three treatment centers had provision to provide free medicines. Association of affordability of medicines with issues of compliance and noncompliance in patients’ of lower socioeconomic status has significant implication in developing countries. However, cost of medications has been found to contribute to noncompliance even in developed countries[9,43] though in a lesser frequency.[13,14] In developed countries, extensive health insurance coverage and financial incentives for long-term treatment may result in relatively better compliance.[53] Lack of time to attend the hospital has been cited as a reason for noncompliance in an earlier Indian study[15] as was the case in the present study. In addition, less frequent follow-up and treatment by proxy were considered to play a significant role in compliance. Earlier Indian studies have observed distance to hospital as one of the reasons for noncompliance.[14,15] However in this study, this factor appeared to play a lesser role in distinguishing between compliance and noncompliance.

CONCLUSIONS

Compliance or noncompliance is determined multidimensionally. Domains related to clinician, family, and medications played a major role in compliance and these have to be reinforced to enhance compliance. Illness-related domain and economic domain played a major role in noncompliance. Specific reasons such as insight into illness and treatment, exacerbation of symptoms on discontinuation of treatment, wish to improve further, good doctor-patient relationship, less side effects, good family support, affordability of medicines, and having time to attend OPD were significantly related to compliance. These have to be addressed to reduce noncompliance. The finding that compliant patients had higher physical comorbidity emphasizes need for detection and adequate management of associated physical morbidity to enhance compliance. Since compliance and thus noncompliance is a dynamic concept, the above-mentioned reasons are to be looked for and addressed continuously during treatment. As total compliance is difficult to achieve, adequate compliance for effective treatment may have to be accepted.

Limitations and implications for future research

This was a cross-sectional assessment. A longitudinal assessment would be more appropriate. The questionnaire used to elicit reasons for compliance and noncompliance was self-designed and not standardized. This study has more relevance in developing countries where economic burden of taking treatment has to be borne by patients and their families and many are ignorant about need for appropriate treatment. Future studies need to be carried out on larger samples and in neurotic spectrum disorders. The setup of service may have some bearing on the compliance and noncompliance, this issue is being addressed in another study.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.