Factors associated with relapse in schizophrenia

Schizophrenia is a chronic and disabling illness that affects approximately 1% of the world’s population. It is often accompanied by relapse even while on treatment.1 Relapse rates vary from 50% to 92%2 and are similar in developed and developing countries, despite the former having well-established mental health services. Among South Africans, there are few published data regarding the prevalence and factors associated with relapses; this study was intended to address that need.


Factors associated with relapse in schizophrenia
Results.Of the 217 patients who were included in the study, 61.8% (N=134) had a history of at least 1 relapse.There was no significant difference (p>0.05) between those who relapsed and those who did not relapse in terms of gender, marital status or employment status.Approximately 46% (N=61) of those who relapsed had co-morbid psychiatric disorders, compared with 10.8% (N=9) in those who did not relapse (p<0.0001),but there was no significant difference between the two groups when comparing the presence of co-morbid medical disorder (p=0.348).Nearly half (N=63) of patients who relapsed had a history of substance abuse (p=0.0054);cannabis was significantly more abused (p=0.0014).Twothirds (N=138) of the study population did not adhere to their treatment, of whom 80.4% (N=107) experienced a relapse (p<0.0001).Significant multiple logistic regression models for patients who relapsed included poor adherence due to sideeffects (odds ratio (OR)=3.032;p=0.023; 95% confidence interval (CI) with schizophrenia.Risk of relapse may be reduced when the treating psychiatrist identifies and addresses these factors.
poly-pharmacology, complex treatment regimens). 13,14Among South Africans, cultural and social attitudes and belief systems are speculated as common reasons for poor adherence to treatment. 15bstance abuse is common among patients with schizophrenia 16 and can lead to relapse independent of its effects on treatment adherence.The lifetime prevalence is estimated to be as high as 47%, 17 with approximately 33% of patients having an alcohol dependence disorder. 18Commonly abused substances include nicotine, alcohol, cannabis and cocaine.In South African studies, alcohol abuse and cannabis abuse were reported as significant factors that contributed to relapse in all mental illnesses, 19 and cannabis, methaqualone and alcohol abuse as having contributed to relapse in acute psychotic states. 20[23] Life stressors may be both internal (e.g.thoughts and feelings) and external (e.g.death of a close relative). 24Other stressors include chronic interpersonal stress, poverty, homelessness, criminal victimisation and stigma.Patients with schizophrenia are more sensitive and more susceptible to the negative effects of even minor stressors.Unemployment and the loss of a close family member are reported as significant causes of relapse in South Africa. 25pression in schizophrenia has been associated with higher rates of relapse, poor outcome, impaired functioning, personal suffering and even suicide. 26,27Approximately 18 -55% of patients with schizophrenia make at least 1 suicide attempt, while 10 -13% of patients succeed in committing suicide. 7e aim of this study was to determine the factors, if any, among patients with schizophrenia that may be associated with relapses.
The specific objectives were to compare the demographic and clinical characteristics of a group of patients with schizophrenia who relapsed, with those who did not relapse.

Method Study sample
The researchers approached the psychiatric nursing staff of mental health outpatient clinics in Johannesburg for a list of possible patients with a diagnosis of schizophrenia.Patients from this list were then randomly selected using the card-shuffling technique.
Patients were included in the study if a review of their records confirmed a diagnosis of schizophrenia according to DSM IV criteria 7 and they had no other psychosis, were ≥18 years old, and had attended the clinics between the period January 1995 and June 2005.Patients were excluded if the diagnosis of schizophrenia had first been made in the preceding 6 months, to avoid any bias in detecting relapses in these patients.
The University of the Witwatersrand's Human Research Ethics Committee (HREC) approved the study.

Relapse and adherence criteria
Relapse was identified in the cases of patients who had documented evidence of either re-emergence or aggravation of psychotic symptoms, a consultation with a psychiatrist and medication change for deterioration of illness, and/or admission to a psychiatric unit in a hospital in accordance with the Mental Health Care Act. 3Planned hospital admission for a non-related illness or for special investigations was not deemed to be a relapse.
Adherence to treatment was considered to be poor if there was failure to fill any prescription, refusal to take medication, stopping treatment prematurely, and reports of taking medication at the wrong time and/or incorrect dosage.
Demographic and clinical characteristics of the patients (gender, age, marital status, source of income, highest level of education, substance abuse, presence and type of co-morbid psychiatric illness, presence of co-morbid medical/surgical illness, stressful life events, presence of and type of stressor, insight) that affect adherence were obtained directly from their case notes.

Statistical analysis
Descriptive statistics were computed as mean and frequencies (count and percentages).The two-tailed paired t-test was used to compare continuous characteristics (age) between the groups.
Multiple logistic regression models were applied to identify factors that best predicted whether or not a patient could relapse. Significant

Relapse rates
The most common outcome of schizophrenia is usually a remitting course with one or multiple relapses in 50 -92% of cases. 2 Patients on medication have a relapse rate of 40%, while those who discontinue their treatment have a 1-year relapse rate of 65% and a 2-year rate of >80%. 28The patients in this study showed similar high rates of relapse.About two-thirds of the patients had at least 1 relapse, with the majority having ≥2 relapses.
In the management of schizophrenia, psychosocial approaches in addition to antipsychotic drug therapy may result in a significant reduction of relapses compared with using antipsychotic treatment only. 28

Partial adherence
In mentally ill patients, partial adherence to treatment remains a therapeutic challenge and a factor that is difficult to quantify; it is compounded by the fact that adherence rates vary for different psychotropic medications: 58% for antipsychotics and 65% for antidepressants. 31[34][35] However, this finding must be qualified in that self-reports of adherence are not always reliable. 31For example, patients might not have revealed the extent of their non-adherence for fear of disappointing the clinician.Nevertheless, other methods of assessing adherence (measurement of medication in the blood or urine, clinician rating of response to treatment, pill count, caregiver reports, presence of medication side-effects) also have limitations regarding reliability.It is important to note, furthermore, that poor adherence is not always significantly associated with relapses. 2e factors associated with poor adherence to treatment in the diagnostic group under review were medication side-effects and lack of insight.Poor insight contributed to a 5.2-times increase in the risk of relapse -a finding that is consistent with other studies. 12wever, other researchers argue that the relationship between insight and adherence is not always straightforward and may be without a direct relationship; 36 e.g. it is possible that the lack of insight in this population might be related to a lower level of formal education and a lack of understanding of mental illness and its phenomena.
Patients who default on their treatment have said that treatment side-effects 14,37 together with complex treatment regimens 38 are the main reasons for their poor adherence.The majority of our patients were on typical antipsychotic medication and reported side-effects (mostly extra-pyramidal effects and impotence) as contributing to their poor adherence, which is similar to other reported findings. 39However, this association between medication side-effects and adherence is not generally clearly established 40,41 because many patients adhere to their medication despite treatment side-effects, 12 and a history of sideeffects is not always predictive of future non-adherence. 36[44][45]

Depressed mood
A quarter of all patients developed a depressed mood during the period studied, increasing the risk of a relapse by 5.3 times.
Other studies on depression associated with schizophrenia show a variation ranging from a high of 75% to a low of 7%. 26,46pression symptoms may appear at any time during the course of schizophrenia and contribute to relapse and a decrease in quality of life. 47,48However, this is contradicted by Robinson et al. 49 who reported that despite a possible relationship between mood symptoms and relapse, neither the severity of baseline depressive symptoms nor the presence of mood symptoms was related to relapse and had no prognostic value.Despite these contradictory findings, early detection and management of depression symptoms may decrease the risk of psychotic relapse and it is therefore important that mental health practitioners should identify depressive moods and separate them from the deficit (negative) features of schizophrenia.

Substance abuse
Although substance abuse was very common among our study population, there was no significant association with relapses.This finding differs from previous findings of strong associations with relapses [50][51][52] and a greater risk of re-hospitalisation. 53Substance abuse may exacerbate psychotic symptoms, and abused drugs may also lead to transient symptom reduction. 54However, some studies report no association between substance abuse and relapse. 55It must be emphasised nevertheless that although it is not clear that substance abuse in patients with schizophrenia results in relapse, it does lead to inter alia increased psychosocial problems, infections, sexually risky behaviour, and hostile and disorganised behaviour -and more so in the presence of other health-related problems, poverty, insufficient food, poor housing and widespread unemployment.

Co-morbid medical illnesses
Surprisingly, only a third of the patients with relapses had comorbid medical illnesses, despite published reports that co-morbid medical disorders exacerbate the relapse process. 56,57In South Africa, mental health services remain marginalised and poorly integrated with general medical services in the primary health care system. 29,49Co-morbid medical illnesses are managed by different services, and it is possible that adequate attention is not paid to recording these co-morbid medical disorders in case notes at mental health clinics.

Age, education, life stressors and patient-provider relationships
Other reported factors associated with relapse include age, 49,58 education, 2 life stressors and patient-provider relationships.
These and other studies months.Chronic schizophrenic patients, furthermore, could be compromised by their illness and unable to establish a committed therapeutic alliance with their psychiatrists.

Limitations
Other reported factors associated with relapses (such as premorbid level of functioning, expressed emotion and duration of untreated psychosis) were not considered in this study and would be better obtained via other study designs.Moreover, the small sample size might have limited our ability to detect statistically meaningful differences.
In any retrospective study design, some data might not have been recorded in case notes, and patients might have been erroneously included or excluded.The design also raises the question of patients' ability to recall events over several weeks between clinic visits, which is compounded by the fact that patients with schizophrenia suffer from some cognitive impairment.However, the majority of our patients' records contained all the data required, and the conclusions drawn are therefore reliable.The study is generalisable, insofar as our study population experiences socio-politico-economic factors and resources that are similar to those experienced by average South African patients with schizophrenia.

Conclusions
Despite recent therapeutic progress, relapse in schizophrenia is a Moosa, MMed (Psych), FCPsych, MCFP F Y Jeenah, MMed (Psych) Division of Psychiatry, University of the Witwatersrand, Johannesburg Aim.Early identification and prevention of relapse in patients with schizophrenia has significant therapeutic and socioeconomic implications.The aim of this study was to determine the factors, if any, that may be associated with relapse in a group of patients in Johannesburg.Method.Patients were recruited from mental health outpatient clinics in a predominantly residential area during the period January 1995 -June 2005.They were included if a review of their records confirmed a diagnosis of schizophrenia according to the Diagnostic and Statistical Manual of Mental Disorders (4th edition) (DSM-IV); they had no other psychotic illness; and they were ≥18 years old.Patients were excluded if the diagnosis of schizophrenia had first been made in the preceding 6 months.Demographic and clinical characteristics of the patients were obtained from their case notes.

Fig. 1 .
Fig. 1.Reasons for poor adherence in the study population.
common and major problem among South Africans.The presence of a co-morbid depressed mood, poor adherence due to a lack of patient insight, and medication side-effects appear to be the factors most likely to increase the risk of a relapse.It is important that, in a local context, the treating psychiatrist identifies and treats mood features along with establishing ways of improving insight and adherence to treatment.If atypical antipsychotic treatments (versus typical antipsychotics) can improve adherence rates, significant reductions in relapse rates and service costs could result.

Table I . Demographic characteristics of the study population
variable) with respect to patient characteristics (age, gender, highest level of education, employment status, poor adherence, substance abuse, co-morbid medical/surgical illness, co-morbid psychiatric illness and psychosocial stressors) were examined by the use of contingency tables (chi-squared test with Yates correction and Fisher's exact test, ORs).The variables showing significant association (p<0.05) in the bivariate analyses were entered into the survey logistic regression to obtain the adjusted ORs.The criterion for removal in the multiple logistic regression analysis was p>0.05.All analyses were calculated by the Statistical Package for Social Sciences 10.0 for Windows (SPSS Inc., Chicago, USA).Volume 14 No. 2 June 2008 -SAJP About 30.6% (41) of patients who relapsed had achieved a primary level of education, and 48.5% (65) a secondary level of education.By comparison, 15.7% (13) of the patients who did not relapse had achieved a primary level of education, and 72.3% (60) a secondary level of education (Fisher's exact=0.006;p=0.01)(TableI).There was no significant difference between the two groups regarding gender (Fisher's exact=1.000,p=0.98);maritalstatus(Fisher'sexact=0.001,p=0.12);receivingstategrants(Fisher's exact=0.471,p=0.52);receivingfamilysupport(Fisher's exact=0.203,p=0.31);andemploymentstatus(Fisher's exact=0.203,p=0.31).Two-thirds (138) of the study population were not fully adherent to their treatment, of whom 80.4% (107) had experienced a relapse (Fisher's exact=0.000,p<0.0001).The factors that were significantly associated with poor adherence included lack of insight (Fisher's exact = 0.000, p<0.0001) and adverse medication effects (Fisher's exact=0.000,p=0003)(Fig.1).Poor relationships with a health care provider (Fisher's exact=1.00,p=0.80)or living great distances from the clinic (Fisher's exact=0.14, p=0.09) did not appear to be associated with relapses.Volume 14 No. 2 June 2008 -SAJP articles *Totals do not add up to 100%.† Respondents in each category who reported 'yes' to having a certain source of income.