Mental health screening questionnaire: A Study on reliability and correlation with perceived stress score

Screening.…


INTRODUCTION
Screening for mental health is essential in clinical practice as well as in research. It plays an important role in detecting mental disorders in primary care set-up. 1 Research has demonstrated that it is possible to construct a self-report questionnaire that can screen mental disorders. 2 3 The importance of screening can be justified by accumulating evidence that diagnostic co-morbidity is under-recognized in routine clinical practice especially in time-and resource-constraint settings. 3 4 The screening instruments offer advantages of ease of administration and economy of professional time, but the validity and practical utility of selfreport screening tests are often questioned. 5 In the context of Nepalese population, the psychometric properties of different scales have been studied in specific groups of population. 6 7 8 9 However, there is dearth of literature on screening of psychological distress among general population of Nepal. This paper describes the development of a 15-item screening questionnaire and testing its psychometric properties for psychological distress in Nepalese population.

MATERIAL AND METHOD
The designed scale is a 15-item instrument to assess psychological distress. The item pool was collected in line with the common psychiatric disorders prevalent in the community. The pool was then discussed among psychiatrist, psychologist, public health experts and community nurse. On the basis of the discussion 15 items were finalized. An additional 16 th question was added which assessed the level of difficulty due to the problems of other 15 questions. The difficulty levels were: not difficult at all, somewhat difficult, very difficult and extremely difficult. Participants responded using a one-point format for Q1-15: 1= Yes, 0 =

ORIGINAL ARTICLE
No. Higher scores indicate higher ratings on the dimension assessed. There is no specific score or cutoff for defining psychological distress in the questionnaire. The distinction is best derived from distributional information from the data collected. For example, high psychological distress could be defined as scores that are in the top 25% (quartile) of the distribution, whereas low psychological distress could be defined as scores that are in the bottom 25% (quartile) of the distribution. Another alternative would be to define high psychological distress as scores that are two standard deviations above the mean, whereas low psychological distress as scores that are two standard deviations below the mean. The response of question 16 could be used as final screening. If a participant has a response of 'very difficult or extremely difficult' they can be advised to seek help of mental health expert considering him/her to be screen positive. The scale was tested among 162 participants from general population after an awareness session about stress and its management, and mental disorders from May 13 th 2019 to July 15 th 2019. The participants of stress management program had to report the problems based on their experience of last 15 days. Along with the scale for psychological distress, perceived stress among the participants was also assessed using "The Perceived Stress Scale (PSS)." 10 It is the most widely used psychological instrument for measuring the perception of stress. It is a measure of the degree to which situations in one's life are appraised as stressful. The questions in the PSS ask about feelings and thoughts during the last month. It has good psychometric properties when used in general population. 11 The question was in English language and only the participants who perceived themselves to be fluent in spoken and written English language were taken. Informed consent was taken from the participants and the PSS score was disclosed to them and based on their score they were advised on stress management and about the need to see a mental health professional. The statistical analysis was done using Easy R (EZR) version 1.40 that is freely available in public domain.

RESULT
Among the participants, 50 were female and 112 were male with mean age of 28.98±8.47 years. The minimum age of the participants was 18 years and maximum age was 69 years. As seen in Table1, scale reliability (Cronbach's alpha) for the designed psychological distress scale was found to be 0.7558 which lies between 0.7 and 0.8, thus having acceptable internal consistency.
The correlation of the designed scale was done with the Perceived Stress Scale. Table 2 shows that the questions of the designed scale had weak to moderate positive correlation with the score on Perceived Stress Scale.

DISCUSSION:
The value of Cronbach alpha of the test scores is 0.7577 which is satisfactory. 12 13 This value is less than the value seen in the study of another screening questionnaire, the Psychiatric Diagnostic Screening Questionnaire (PDSQ) which had overall reliability of 0.85. 14 Similarly, another screening tool, Brief Symptom Inventory had the reliability of 0.75 to 0.85 for different diagnoses that is comparable to our study. 15 Internal consistency coefficients of the Brief Symptom Rating Scale (BSRS-5) ranged from 0.77 to 0.90 which is also comparable to our study. 16 When compared to other screening tools, our scale has less number of items and hence is more convenient to use with comparable internal consistency. As described earlier it has 15 questions to look at the psychological symptoms and the 16 th question assesses whether the symptoms have caused difficulty in day to day activities (graded into four options: not difficult at all, somewhat difficult, very difficult and extremely difficult). The scale doesn't give the cut-off score from the first 15 questions. However, depending upon the response of 16 th question, an approximate estimate of the problem can be made. This is in line with our concept of dysfunction as the core of mental disorders. 17 18 The items of the designed questionnaire showed weak to moderate positive correlation with the scores on Perceived Stress Scale. It is a known fact that perceived stress influences the mental health status and psychological well-being. 19 20 There is a direct association between perceived stress and psychological distress and mental illness. It is seen that the scores on Perceived Stress Scale and its factors correlate significantly with depression scores. 21 In the large, population-based study, it has been seen that middle-aged men and women who experienced depression or anxiety had more perceived stress. 22 Similarly, the patients with Obsessive-Compulsive Disorder (OCD) also demonstrate a higher level of perceived stress than general population. 23 In a study of 360 undergraduate students, it was seen that perceived stress scores had positive correlation with suicidal ideation. 24 There are other robust evidences showing higher perceived stress predicts greater odds of suicidal ideation. 25 In cases of psychosis, large epidemiological studies have found greater stress sensitivity and perceived stress to be associated with increased odds of psychotic experiences. 26 In the case of mania, it would be difficult for a patient to acknowledge the stress, hence our study shows a very weak positive correlation between the symptom of mania and perceived stress score.
There are reports suggesting substance use disorders are also related to higher perceived stress score. 27 28 The difficulty level the symptoms have created as reported in Q16 also shows moderate positive correlation with the stress score. Hence, pointing to these indirect evidences, we conclude that the newly developed mental health screening questionnaire can be used to measure the psychological distress in general population. The major strength of the study is that it is one of the few studies in Nepalese general population regarding psychological distress and mental health problems. The scale has less items and therefore, is less time consuming and is selfreported. The positive correlation shown with the already validated Perceived Stress Scale is another strength. The age of the sample is between 18 years to 69 years, so some amount of generalizability for age is present. However, there are some limitations to our work, as this is only a screening questionnaire, therefore, no diagnosis can be made based on this scale and in screening not all mental health conditions can be assessed.
Like all other self-reporting instruments, this questionnaire cannot be free of biases. Our study doesn't account for various confounders affecting the total score. The questions are also not validated with gold standard and scale's diagnostic accuracy has not been studied. Similarly, in the second part of the study we have correlated the items of the developed scale with the score on Perceived Stress Scale and concluded that we might be able to screen psychological distress and this might not be true in severe mental disorders with no insight like mania or psychosis. Hence, we recommend the sensitivity, specificity, and diagnostic accuracy to be measured in another study. We also recommend the scale be translated in Nepali language with stratified sampling so that it can be used in heterogeneous population to look for generalizability.

CONCLUSION:
Despite many shortcomings of our scale we may be able to use it for basic screening of psychological distress. The scale has acceptable internal consistency and the items are positively correlated with the scores on Perceived Stress Scale. Further modifications and studies in larger heterogeneous population for validity are warranted.