The study was a prospective study which was conducted after getting approval of the hospital ethics committee. The study period was for six months from January of 2009 till June 2009. This study was based on subjects who had been part of a larger study done by the same authors which has been published earlier which looked at quality of life relationship to psychopathology severity [4]. The sample size was calculated with a power of 0.8 and confidence interval of 95% also limited by the time duration available for the study which was short.

The sample comprised 40 subjects who had satisfied the inclusion criteria: (a) patients aged 18 years and above; (b) fulfilling ICD-10 criteria for schizophrenia [5]; and (c) having an acute exacerbation defined as Clinical Global Impression Severity (CGI-S) scale score of 4 or more which necessitated admission in hospital. Subjects who refused to give written informed consent were excluded from the study [6]. The subjects were recruited from the patients admitted in Kasturba Medical College, Manipal, a tertiary hospital in south India under Department of Psychiatry. A convenience sampling method was used. The diagnosis was confirmed by consultant psychiatrists using ICD-10 Diagnostic Criteria for Research (ICD-10 DCR) [5].

Data was collected on the basis of assessment at admission, discharge from hospital and three months after first assessment. The diagnosis were ascertained by face to face clinical interview of the patients and their informants by the psychiatrists using the ICD-10 DCR criteria. All consenting adults who fulfilled the inclusion criteria were first administered the socio-demographic proforma by the investigator. The Mini International Neuropsychiatric Interview (MINI) which is a short structured interview was used to confirm the diagnosis and rule out other psychiatric disorders [6]. All the patients in the study received treatment as per individual consultant’s decision which included pharmacological management with antipsychotics (first generation or second generation) and psychosocial management including psycho-education and family therapy. The severities of symptoms as assessed by clinician were rated at admission on CGI-S scale and improvement was rated on Clinical Global Impression- Improvement (CGI-I) scale at discharge and at end of 3 months [7]. The Positive And Negative Syndrome Scale (PANSS) and Schedule For Assessment for Psychiatric Disability (SAPD) scale were used for assessment at baseline and at end of 3 months [8,9]. All the instruments used were used obtained from the online sources of the same freely usable with purpose of doing research as a student.

In addition to the total PANSS score (sum of all 30 items, maximum score 7 per item), five sub-scales were calculated [10]. These sub-scales were:

The SAPD was used to assess disability in social functioning. The instrument was developed from the Disability Assessment Schedule of the World Health Organization and was standardized for use in India [9]. Information given by both the patient and the key informant formed the basis for scoring on the instrument. They were asked to use the last one month functioning of the patient as reference period. The SAPD scored disability into three domains of functioning which are general activity, social role and occupational role.

Analysis of data was carried out using SPSS software package (version 10, SPSS Inc, Chicago, USA). Descriptive statistics were used. Parametric tests, the paired t-test, pearsons’ correlation test were used to look for changes in severity of symptoms and disability. Co-relation of statistics was used to assess relations between symptom severity and disability.

Out of the 40 subjects who were included in the study only 37 completed the study and the other 3 were lost during follow up. To prevent skewing of data towards subjects who might have had better response to medications. As it was probable that subjects who dropped out early were the ones who might have not had much improvement. Therefore, for baseline their data was also used but to correlate, if change of symptoms had association with disability only subjects whose full data was available was used.

[Table/Fig-1] presents the socio-demographic distribution of the study population (n=40), the mean age of 34.27 (SD±9.72), males were the majority 23 (57.5%), most of the subjects were having a high school education 17 (42.5%), most of the subjects were single 21 (52.5%), 14(35%) of the subjects were homemakers, the family income per month showed that the most frequent range being 2000-5000 having 13(32.5%) of the subjects.

[Table/Fig-2] presents the illness variables and it showed that mean duration of illness was 7.92 years (SD±6.738), the number of subjects having known medical co-morbidity were 8 (20%) most of which were cardiovascular or endocrinal. There were 28 (70%) subjects who had other co-morbid psychiatric disorders along with diagnosis of schizophrenia mostly they had substance abuse disorders. The severity of illness as indicated by CGI-S showed that 18 (45%) subjects had severe illness, the CGI-I which indicated overall improvement of illness for the 37 subjects assessed at endpoint showed that 27 (67.5%) subjects had much improved from baseline.

[Table/Fig-3] presents the results of paired t tests done for the various symptoms domains assessed by PANSS (higher scores indicate increased severity of illness) at baseline and at endpoint for the 37 subjects who completed the study. It showed that except for the affective (anxiety/depression) domain (mean=0.486, t=0.937, p=0.35), the other domains of PANSS showed significant change maximum was for positive domain (mean=12.054, t=14.214, p<0.01).

[Table/Fig-4] presents the results of paired t-tests done for the disability domains as assessed by SAPD (higher scores indicate increased disability) at baseline and at endpoint for the 37 subjects who completed the study. It showed that all of them were showing significant change maximum was for behavioural functioning score (mean=7.783, t=12.337, p<0.01).

[Table/Fig-5] presents the correlation of baseline PANSS total scores and domain scores to the various baseline disability domains. It showed total scores, negative domain score and disorganization had significant positive correlation of total scores to behavioural functioning and occupational functioning of the disability domains baseline PANSS excitement domain score showed significant positive correlation to the social functioning (r=0.417, p=0.007) among the disability domain scores. No change in significant values even if the 3 subjects lost to follow-up was removed from the data.

[Table/Fig-6] presents the correlation of change in total PANSS score and domain scores from baseline to the endpoint for the various endpoint disability domains. It showed total score, change in positive domain score, and change in negative domain score showed significant negative correlation to all disability domain scores. Change in disorganization domain score showed significant negative correlation to only occupational functioning domain score. Change in excitement domain score showed significant negative correlation to behavioural functioning score and social functioning of disability domain scores.

The present study made an attempt to determine whether change in severity of symptoms was having any relationship to the person’s disability. As most of the subjects were single so many of the questions in part 3 of the SAPD was not applicable and as was done in studies using this scale earlier also while calculating scores for the whole sample they were left out for all of them [9].

The study centre caters to a predominantly rural/suburban catchment area. So a majority of subjects were from the rural/suburban area. Average age was around 34 years with males being slightly higher in number; this was similar to findings in other studies [11,12]. It has been considered that being married indicates good outcome and in Indian context most studies have found individuals are married. Most of the subjects were of low to middle socioeconomic status and were mostly labourers or homemakers as in other studies [13]. Most of them had a high school education or less as in other studies, but some studies had found that level of education was higher and it has been considered that higher level of education will positively affect later functioning [14,15].

The analysis of the illness variables revealed that mean duration of illness was 7.92 years this is similar to other studies [14,15], but other studies have taken subjects with more than 10 years of illness [12,13]. Longer duration of illness at intake has been found to have a negative effect on overall symptoms at follow-up. A 20% of the subjects had a known medical co-morbidity, most common diagnosis were diabetes mellitus, dyslipidemia and hypertension. Possible cause for this may be that these patients are on psychiatric drugs already. Around 70% of the subjects had a co-morbid psychiatric diagnosis, this was in most cases substance abuse disorders predominantly tobacco, this was similar to other studies [13]. As this study had a requirement for acute exacerbation of schizophrenia needing hospitalization as an inclusion criteria, majority of the subjects who were enrolled in the study had a moderate to severe intensity of the illness as per the CGI-S scale which is similar to other studies [10]. At the end of study period of the 37 subjects who completed the study only 17.5% of them had shown minimal improvement while the rest had shown much or very much improvement in their clinical symptoms as per the CGI-I scale. It had been found in a meta-analysis of outcome that around 45% have good clinical outcome at 1 year [16]. As this study was of short duration of three months so outcome may be different from published findings.

Change in psychopathology from baseline to endpoint was examined and it was seen that there was a significant decrease in the total PANSS score. Also, looking at the various domains of the PANSS it was seen that except for the affective domain which did not show a significant change, the other four had shown a significant decrease with the maximum being seen in the positive domain and the least being in negative domain. Karow A et al., had found that there was a significant change in all domains and total PANSS score also [17]. This investigator’s findings are similar to those by Ho BC et al., in their study [18]. Weiselgren IM et al., has suggested that symptoms in acute recent onset patients may have better predictive validity than those ascertained during an acute exacerbation in patients who have been ill longer [19]. The present study had a mixed population of first episode and chronic schizophrenics which may have affected findings.

SAPD scale which was used to assess disability is a modification of the WHODAS II and has been found to be useful in assessing patients of schizophrenia from developing countries [9]. Other studies have used Global Assessment of Functioning (GAF) scale, Social and Occupational Functioning Assessment Scale (SOFAS), Social Functioning Scale (SFS) [20]. The SAPD gives higher score when disability increases. Change from baseline to endpoint was assessed and it was found that significant reduction in all the domains of the disability as well as overall disability were found, this was similar to other studies [13]. The SAPD gives higher score when disability increases. Change from baseline to endpoint was assessed and it was found that significant reduction in all the domains of the disability as well as overall disability were found, this was similar to other studies. On the other hand, the presence of a good support system of family members might be a possible explanation for the significant reduction in disability seen here and as reported in the previous Indian studies.

When looking at the scores at baseline when the subject is in acute exacerbation it was seen that total PANSS score had a significant positive correlation to the behavioural and occupational functioning. This result is in the expected direction. In the present study the analysis of the disability domains showed that PANSS positive and affective domains had no significant correlation. Possible explanation for this is that the patients are hospitalized during this period and as florid psychotic symptoms are expected in them thus they are less disruptive to the person’s ability to meet their needs. While the PANSS negative and disorganization domains had significant positive correlation to the behavioural, occupational and overall disability, the excitement domain was having significant correlation to only the social functioning score. Thirtahalli J et al., had found that severity of symptoms was correlated to increase in disability and the present study reflects a somewhat similar finding [13].

The analysis showed that change in total PANSS score had significant negative correlation to all the disability domains. On analysing the relationship of change in PANSS domain scores to the disability domains, it showed that change in PANSS positive, negative domain had negative correlation to all the disability domains. The change in PANSS disorganization domain had negative correlation to only the occupational functioning, the change in PANSS excitement domain had negative correlation to the behavioural and social functioning domains and finally the change in PANSS affective domain had no significant correlation to any of the domains. Thirtahalli J et al., had found that change in PANSS negative domain had the strongest correlation to disability [13]. Saraswat N et al., had found that the positive and negative domains are significantly correlated to disability [20]. Seigel SJ et al., had found that both change in PANSS positive and negative domain had strong correlation to disability also they felt it is the type of symptom patient has than the intensity of the symptoms which determine the later functioning [15]. Akinsulore A et al., had found that positive, negative and affective domains have significant correlation to the disability domains, they felt it was due to poor social skills, lack of motivation and self stigma [21]. The better functioning due to fall of PANSS negative score is possibly due to the fact that some of the initial score was possibly due to secondary negative symptoms which resolved on treatment. Ritsner M et al., has also explained that the improvement in social relations is related to negative symptoms and social support as good family support helps improve the social relations and hence may effect disability [22].

Limitations of this study were several. The sample size was small so the strength of correlation seen here cannot be generalized. There was no comparison group to look at whether the treatment was the cause for change. The duration of follow-up was short so they may not be able to detect substantial changes. The sample was not homogenous, it included both first episode and chronic patients which could have made interpretation of findings difficult. The effect of type of treatment given was not taken into consideration as this may also possibly affect the endpoint variables.

This study adds to the evidence which is increasingly showing that we need to look at schizophrenia as not purely as a psychotic disorder but rather a multifaceted one which requires interventions which would benefit in improving the persons social functioning ability also.

This study found that disability had a positive correlation to the negative, excitement and disorganization domains of PANSS. Correlation of change in PANSS domains from baseline to endpoint with change in disability had a negative correlation except for the affective domain.

Further research with larger samples and more homogenous population is needed to identify the strength of the associations of the domains of schizophrenia with domains of disability.