Development and validation of the NIMHANS-NMHS-2 disability scale (NNDS)

Abstract

Background:

Measuring disability is one of the primary objectives of India’s National Mental Health Survey-2 (NMHS-2). The survey required a scale that was short, free to use, culturally flexible, administrable by lay-interviewers, and measuring disability aligned with the current understanding of mental disorders.

Aim:

Lacking a scale meeting these requisites, the NIMHANS-NMHS-2 Disability Scale (NNDS) was developed as a new adult disability scale. This manuscript reports on its development and validation.

Methods:

Nineteen items across six domains were generated by desk review. Content validation conducted by 16 Consultants at the National Institute of Mental Health and Neuro Sciences led to removal of six items (Content Validity Index <0.75). Analysis included Cronbach’s alpha for internal consistency and Exploratory Factor Analysis. K-means cluster analysis identified disability severity groups, while receiver operating characteristic analysis determined optimal disability cut-offs based on WHO-Disability Assessment Schedule (WHODAS) and NNDS scores.

Results:

The NNDS was validated across inpatient (n = 141) and Outpatient (n = 142) sample from NIMHANS, and a community sample (n = 133). It demonstrated excellent internal consistency (α-0.94), a stable five-factor structure, and strong correlations with the WHODAS 2.0 (rho = 0.93), and Brief Psychiatric Rating Scale (rho = 0.80). Cut-off scores established were: No disability: 13–17; Mild: 18–31; Moderate: 32–40, and Severe: >40.

Conclusion:

The NNDS has demonstrated robust psychometric properties for measuring disability. Being open-source, the scale has broad applicability for clinical practice, community assessments and research. (For the NMHS-2, an eight-item short version is being used, its development and validation are separately submitted for consideration of publication).

INTRODUCTION

The National Mental Health Survey-2 (NMHS-2) (2024–26) of India is currently underway across all the 28 states and 8 union territories of the country. The survey has two primary objectives, the first is to examine the prevalence of mental health morbidity and the second to examine the disability due to mental health conditions. The NMHS-2 uses the Flexible Interview of International Classification of Diseases 11 (ICD-11) (FLII-11), a World Health Organization (WHO) structured diagnostic interview (SDI), validated independently to assess mental health morbidity for the NMHS-2.[1,2]

Mental illnesses contribute to significant disability, with depression, anxiety disorders, schizophrenia, and substance use disorders among the leading causes of disability in the world.[3] However, these inferences are mostly from studies on clinical samples where the estimation of disability is likely to be higher owing to greater severity of illness. Disability estimates in community surveys tend to be more representative of the general population in a country. Additionally, quantifying disability will further help us understanding the relationship between reported prevalence and real-world impact.[4,5] Furthermore, the precise measurement of the burden of mental illness influences mental health policies and resource allocation. Hence for NMHS-2, measuring disability linked to mental health disorders was deemed an equally important objective of public health importance.

The NMHS-1 used the 3-item Sheehan Disability Scale, a brief tool, now a paid resource.[6] For the NMHS-2, it was determined that a more comprehensive disability examination would be better to quantify the full extent of the burden of mental illness. Currently in India, the Indian Disability Evaluation and Assessment Scale (IDEAS) and the WHO-Disability Assessment Schedule-12 item (WHODAS 2.0) are the most used tools for disability assessment. The IDEAS has been developed and validated by the Indian Psychiatric Society and subsequently adopted by Ministry of Social Justice and Empowerment to provide a standardized assessment for disability in those with severe mental health disorders.[7] However, its limited application to less severe mental health conditions and community settings[8] is a limitation. The WHO-Disability assessment scale (WHO-DAS 2.0) is also widely used and has the advantage of being used both in hospital and community settings.[9] Nonetheless, the instrument is generic and several WHO-DAS 2.0 items are deemed unrelated to mental health and was not considered both for the NMHS-1 and NMHS-2.[10] Other scales like the schedule for assessment for psychiatric disability (SPAD), USA Social Security Scale, the global assessment of functioning (GAF) had several constraints for use in India. These scales had several limitations: they were not suitable for large-scale community surveys, required cultural adaptation and local validation, and needed rater with mental health expertise.

Given an absence of an appropriate instrument for evaluating disability within the requisites of the NMHS-2, the central co-ordination team of the NMHS-2 at the National Institute of Mental Health and Neuro Sciences (NIMHANS), Bengaluru, undertook the development of a new disability scale for adults - NIMHANS-NMHS-2 Disability Scale (NNDS). The scale had the objective to be brief, easy to administer for lay-interviewers, culturally sensitive to the country’s diversity, and evaluating disability domains that are consistent with our present knowledge of mental diseases. This paper presents development and validation of NNDS Scale.

METHODS

Developing the theoretical framework for the NNDS

Psychiatrists from the NMHS-2 central team initially conducted a desk review to examine the common constructs for the measurement of disability. The review’s focus was on the various assessment techniques used in clinical and community-based surveys, the domains affected, and the current understanding of disability in mental health illnesses. A team of three senior mental health professionals from NIMHANS examined the findings and reached a consensus on a three-point theoretical framework for development of the disability scale.

Three key areas of the frameworks were:

1. Domains of assessment: A comprehensive disability assessment must be designed to cover the six major life domains impacted by mental health disorders: self-care, daily functioning, learning/cognition, interpersonal functioning, work, and participation (in home and social settings).

2. Time required for assessment: In the NMHS-2, disability was only one of the many assessments to be conducted. Therefore, it was imperative that the newly developed scale achieve a balance between comprehensiveness and brevity; it had to include all essential domains of disability, while being brief enough to prevent participant fatigue and inconvenience.

3. Evaluation period: Disability assessments typically focus on either a patient’s worst period of illness or their current level of functioning, with both most evaluated over a one-month (30 day) period. The expert group evaluated both options and determined that for the NMHS-2, assessment of current disability over a 1-month period would be more effective in preventing recall bias and yielding more meaningful data.

Item generation

Items were generated for each domain using a combination of both deductive and inductive reasoning. This dual approach was to ensure that the items were not only theoretically sound (deductive), but also reflect real-world experience (inductive), resulting in a comprehensive and valid set. Based on the defined theoretical framework, a total of 19 items covering the six domains of disability were created to build the first draft disability scale. For each item, there were three suggested formats for the scoring instructions.

Cognitive interviews

The draft disability scale was then administered to English-speaking patients and their caregivers by the project psychiatrist. This was specifically to explore the thought processes underlying their responses to each item, identifying difficult to answer questions and items requiring more linguistic clarity to improve the assessment. Response saturation was noted after eight interviews. Following which, there were minor modifications in three questions.

Content validation

Content Validation of each item of the draft disability scale was done by a panel of 16 experts from NIMHANS. The panel comprised consultants from the Departments of Psychiatry (n = 8), Clinical Psychology (n = 4), and Psychiatric Social Work (n = 4), all of them with extensive experience in psychiatric rehabilitation.

The expert panel opined on the items, the proposed scoring format for the scale and applicability of each item to the NMHS-2. Item evaluation was based on four criteria: ease of understanding, relevance, subjectivity, and utility. A Likert scale with scores 1–4, was used for rating, with higher ratings indicating a better fit.

The Item level Content Validity Index (I-CVI) was calculated for each item. Items with a I-CVI below the 0.75 cutoff[11] were removed. Due to which, five items were removed. The experts’ suggestions for potential revisions were then reviewed, and two overlapping questions were combined into one. The scoring approach with highest CVI (0.82) was selected.

The final 13 items NIMHANS-NMHS-2 disability scale and the selected scoring approach subsequently underwent clinical validation [Table 1].

Table 1.

NIMHANS- NMHS2 Disability Scale

Information Sheet
This is a scale for measuring disability for psychiatric disorders both in the clinical and general population. The instrument has been developed for the National Mental Health Survey-2 (India). The major domains assessed in this scale are: Self-Care (Items 1-2) Daily functioning (Items 3-4) Learning/cognition (Items 5-7) Interpersonal functioning (Items 8-9) Participation (Items 10-11) Work (Items 12-13) The scoring for the questions is to be done as follows: 0 = No difficulty (means the person has no problem) 1 = Mild difficulty (means a problem that is present less than 25% of the time, with an intensity a person can tolerate and which happens rarely over the past one month) 2 = Moderate difficulty (means a problem that is present less than 50% of the time, with an intensity, which is interfering in the person’s day-to-day life and which happens occasionally over the past one month) 3 = Severe difficulty (means a problem that is present more than 50% of the time, with an intensity, which is partially disrupting the person’s day-to-day life and which happens frequently over the past one month) 4 = Profound difficulty (means a problem that is present more than 95% of the time, with an intensity, which is totally disrupting the person’s day-to-day life and which happens every day over the past one month) 8 = Not specified (means there is insufficient information to specify the severity of the difficulty) 9 = Not applicable (means question is inappropriate)
NIMHANS- NMHS2 Disability Scale
Instructions The questionnaire is to be administered by a trained interviewer and examines difficulties due to mental health issues. The respondent is required to think about the difficulty experienced in the following activities in the past one month. Information from a close informant may be required to answer question 6 and 7 (with the respondent’s consent).
For each question circle only ONE response.
Items	No difficulty (0)	Mild difficulty (1)	Moderate difficulty (2)	Severe difficulty (3)	Profound difficulty (4)	Not specified (8)	Not applicable (9)
1. In the last one month, did you have difficulty in taking care of your daily activities (e.g.: brushing, bathing, eating, dressing)?	0	1	2	3	4	8	9
2. In the last one month, did you require any assistance to take care of yourself (e.g.: bathing, eating, dressing)?	0	1	2	3	4	8	9
3. In the last one month, did you have any difficulty in carrying your routine household chores (washing, shopping, cleaning, cooking etc.)? (please consider only activities routinely done when well)	0	1	2	3	4	8	9
4. In the last one month, did you have any difficulty in carrying out for your day-to-day tasks within the reasonable time expected (washing, cleaning, cooking etc.)? (please consider only activities routinely done when well)	0	1	2	3	4	8	9
5. In the last one month, how much difficulty did you have in concentrating on things (ex. forgetting what was being done, loosing track of conversation)?	0	1	2	3	4	8	9
Domains	No difficulty (0)	Mild difficulty (1)	Moderate difficulty (2)	Severe difficulty (3)	Profound difficulty (4)	Not specified (8)	Not applicable (9)
6. In the last one month, did you experience any difficulty in understanding daily information (e.g., understanding instructions for crossing road, playing cards, making tea etc.)?	0	1	2	3	4	8	9
7. In the last one month, did you have any difficulty in remembering things (ex. whom you met recently, what you ate in last meal)?	0	1	2	3	4	8	9
8. In the last one month, did you experience any difficulty in your communication with people (e.g. greeting, talking)?	0	1	2	3	4	8	9
9. In the last one month, did you have any difficulty in maintaining relationships with those whom you are close to (e.g., friends/spouse/ siblings)?	0	1	2	3	4	8	9
10. In the last one month, did you have any difficulty in involving yourself or participating in community activities (e.g. attending social gatherings/ marriage functions/religious activities)?	0	1	2	3	4	8	9
11. In the last one month, did you have trouble in involving yourself with others in decision making pertaining to life affairs (planning a trip, buying an appliance, making investments)?	0	1	2	3	4	8	9
12. In the last one month, did you have any difficulty in performing your regular and expected work (academics/household/occupational work)?	0	1	2	3	4	8	9
13. In the last one month, did you have any difficulty in carrying out a task that you know how to do (e.g.: cooking, driving, money transactions)?	0	1	2	3	4	8	9

^@The scoring for Question 2 is the same as the other questions; however, the term “assistance” replaces “difficulty” in the scoring instructions

Validation in clinical and community settings

The NNDS was administered in clinical and community settings to establish both its concurrent validity and discriminant validity.

Settings

The study was conducted among three groups of English-speaking adult participants selected through convenience sampling. Two groups were patients from the out-patient or in-patient services of the Department of Psychiatry at NIMHANS. The third group were the subjects from the community, individuals from three the urban wards in the immediate vicinity of NIMHANS. Participants were assessed by trained rater’s who had post-graduate qualification in psychology or social work.

Process

Consenting participants were administered the NNDS, along with a study specific proforma which recorded basic socio-demographic information. For patients, the clinical diagnosis and the recruitment setting (out-patient/in-patient) were recorded.

The concurrent and discriminant validity of the NNDS was established by using the following structured instruments:

WHO-disability assessment schedule (WHODAS 2.0)

The 12-item WHODAS 2.0 was used as the gold standard to establish the concurrent validity of the NNDS. The WHODAS 2.0 is a widely used generic instrument to measure disability, and functional impairment for chronic medical and mental conditions. It has been developed through extensive cross-cultural studies, including in India.[9,12]

Brief psychiatric rating scale (BPRS)

The 18 items—Brief psychiatric rating scale (BPRS) was used to examine whether the NNDS could differentiate participants with varying severity of illness, and recruitment settings. The BPRS has been conceived to measure psychopathology and treatment effectiveness.[13,14] It assesses for a comprehensive set of common symptoms and has been used as a transdiagnostic instrument.[15]

Inter-rater reliability

To test inter-rater reliability, a second rater was present and independently rated the first rater’s assessment (joint inter-rater reliability).

Sample size calculation

The minimum required sample size for the clinical validation of the NIMHANS-NMHS-2 disability scale was estimated to be 300. This was based on presuming a minimum correlation of 0.75 between the NNDS and the WHODAS-12, with a 95% confidence interval and ± 0.5 margin of error. An attempt was made to recruit equal number of participants from out-patient (OP), in patient (IP), and the community.

Statistical analysis

Statistical Package for the Social Science (SPSS) Version 28[16] was used for all analyses. Descriptive statistics, including means, standard deviations, medians, inter-quartile ranges, and percentages were used to summarize demographic and clinical characteristics of the three study groups: in-patients (IP), out-patients (OP), and community sample.

The internal consistency of the 13-items of NNDS was measured by the Cronbach’s alpha. Exploratory factor analysis (EFA) using the Spearman rank correlation was conducted to identify the scale domains. The data-set showed excellent suitability for EFA, evidenced by a Kaiser-Meyer-Olkin (KMO) value of 0.94 and a significant Bartlett’s Test of Sphericity (P < 0.001). Parallel analysis was conducted using 1,000 simulated datasets to determine the optimal number of factors. Eigen values were estimated using squared multiple correlations, applying both the average of simulated eigen value criteria and the 95^th percentile criteria suggested by Crawford et al. (2011).[17] The results suggested retaining models ranging from four to six factors using generalized least squares extraction method followed by promax and oblimin rotations.

The Spearman’s rank correlation coefficient was used to evaluate the strength of relationship between the three clinical scales. The joint inter-rater reliability was estimated using two-way mixed model with single-measure absolute agreement intraclass correlation coefficient (ICC).

To identify subgroups based on disability scores, K-means cluster analysis was conducted using WHODAS and NNDS scores. The number of sub-groups (none, mild, moderate, and severe disability) (k = 4) was selected based on interpretability and clinical relevance. Three separate receiver operating characteristic (ROC) analyses were performed to determine optimal cutoff scores on the NNDS for categorizing the four groups (none vs mild, mild vs moderate, and moderate vs severe). The area under the curve (AUC) was used as a measure of classification performance. Optimal cut-off scores for categorizing levels of disability were determined using the maximum Kolmogorov-Smirnov (K-S) statistic method.[18]

RESULTS

A total of 416 adult participants were recruited for the study, with approximately equal representation across the three groups: IP (n = 141), OP (n = 142), and Community sample (n = 133). Both patient groups had a higher proportion of males, reflective of gender distribution within the hospital. The higher proportion of females in the community sample is likely due to their greater availability at home during working hours, with males being mostly at the workplace [Table 2].

Table 2.

Socio-demographic and clinical characteristics of the study sample

Variables	In-Patient (IP) n=141	Out-Patient (OP) n=142	Community n=133
Age (years) (Mean±SD)	33.7±12.1	34.4±10.7	41.1±15.9
Gender (Male %)	93 (71.0%)	107 (70.4%)	58 (43.9%)
BPRS Score
Median     (Q1, Q3);     Min-Max	50.0 (43.0, 59.0) 36–93	44.0 (39.0, 50.0) 36–70	36.0 (36.0, 36.0) 36–64
NNDS Disability Scale Scores
Median     (Q1, Q3);     Min-max	32.0 (22.0, 37.0) 13–52	20.5 (14.0, 29.5) 13–50	13.0 (13.0, 13.0) 13–37
WHODAS Scores
Median     (Q1, Q3)     Min-max	29.0 (19.0, 35.0) 12–48	19.5 (14.0, 28.8) 12–46	12.0 (12.0, 12.0) 11–31
Diagnostic Categories	IP n=141 (%)	OP n=142 (%)	Total n=283 (%)
Schizophrenia	58 (41.4)	16 (11.2)	74 (26.1)
Depressive disorders	16 (11.4)	18 (12.7)	34 (12.0)
Bipolar Affective Disorder	25 (17.8)	14 (9.8)	39 (13.8)
Disorders due to substance use	8 (5.7)	57 (40.2)	65 (23.1)
Obsessive compulsive disorder	21 (14.9)	11 (7.7)	32 (11.3)
Anxiety and fear-related disorders	1 (0.1)	20 (14.2)	21 (7.4)
Others (Gambling and gaming disorders, ADHD)	0 (0.0)	4 (2.8)	4 (1.4)
Secondary Mental disorders	12 (8.7)	2 (1.4)	14 (4.9)

Min-Max: Minimum – Maximum. NNDS=NIMHANS-NMHS-2 Disability Scale, BPRS=Brief psychiatric rating scale, WHODAS=WHO-Disability Assessment Schedule

The scores for the BPRS, NNDS and WHODAS-12 followed a consistent gradient, being the highest in IP, intermediate in OP, and lowest in Community sample [Table 2].

The patient sample represented six major psychiatric diagnostic categories as well as secondary mental disorders, and their relative proportions reflected the overall pattern of care-seeking at the hospital [Table 2]. For patients, formal diagnostic assessments were not conducted at intake; instead, clinical diagnoses recorded in patient files following a detailed assessment by a Consultant Psychiatrist was considered. Community participants underwent brief clinical screening by project psychiatrists to identify the presence of diagnosable mental health disorders.

The internal consistency of the 13 items of NNDS, measured by the Cronbach’s alpha, was 0.94. Domain specific values were not calculated due to limited number of items in each domain.

The EFA with oblimin rotation results indicated that the NNDS was best represented by a five- factor model comprising the following domains- self-care, daily functioning, learning and cognition, interpersonal functioning/participation, and participation/work. The fit of the five-factor solution was evaluated using several standard goodness-of-fit indices and the five factors explains 72% of the variability in the results. The analysis yielded a Tucker-Lewis Index (TLI) of 0.961 and a Root Mean Square Error of Approximation (RMSEA) of 0.069 [Table 3].

Table 3.

Five-factor model solution for NIMHANS-NMHS2 disability scale

Item	Learning and cognition	Daily functioning	Selfcare	Interpersonal functioning	Participation/work
Selfcare: Question 1 (Daily activities)	0.03	0.18	0.72	0.03	−0.01
Selfcare: Question 2 (Assistance in daily activities)	−0.02	−0.04	0.88	−0.01	0.04
Daily functioning: Question 1 (Routine household chores)	0.03	0.85	0.06	0.08	−0.06
Daily functioning: Question 2 (Time taken for chores)	0.05	0.82	0.01	−0.04	0.10
Learning and cognition: Question 1 (Attention)	0.87	0.06	−0.04	−0.03	0.05
Learning and cognition: Question 2 (Comprehension)	0.47	0.00	0.14	0.22	0.05
Learning and cognition: Question 3 (Recent memory)	0.70	0.03	0.06	0.05	0.00
Interpersonal functioning: Question 1 (Communication)	0.11	0.11	0.11	0.65	−0.01
Interpersonal functioning: Question 2 (Maintaining relationships)	0.34	−0.09	0.17	0.44	0.04
Participation: Question 1 (Community activities)	−0.06	0.22	0.01	0.51	0.32
Participation: Question 2 (Involving in decision making)	0.17	0.04	0.03	0.30	0.39
Work: Question 1 (Occupational functioning)	0.12	0.17	0.11	0.00	0.58
Work: Question 2 (Learned task difficulty)	0.19	0.00	0.20	0.07	0.48
Sum of squared loadings (Eigen values of the common factor structure)	2.35	1.98	1.92	1.68	1.43
Percentage variance explained	18	15	15	13	11

The NNDS showed a very strong correlation with the gold standard, WHO-DAS 2.0 (rho = 0.935). This strong correlation held true for both out-patients (rho = 0.89) and in-patients (rho = 0.87) when examined separately [Table 4]. The joint inter-rater reliability with 31 participants for the NNDS was almost perfect with ICC value of 0.998 (95% confidence interval of 0.996, 0.999]. Additionally, the NNDS also demonstrated a very strong correlation with the structured measure of psychopathology, the BPRS, across the entire participant group (rho = 0.80). In sub-group analysis, the correlation with BPRS was strong for out-patients, moderate for in-patients and no correlation for the community sample [Table 4].

Table 4.

Spearman’s rank Correlations and corresponding 95% confidence interval of the NIMHANS-NMHS-2 disability scale with BPRS Score and WHODAS Scores

Correlation	Spearman’s rank correlation (95% Confidence Interval)
	All participants (n=416)	In patients (n=141)	Out patient (n=142)	Community (n=133)
NNDS Score - WHODAS Score	0.93 (0.91, 0.94)	0.87 (0.82, 0.91)	0.89 (0.85, 0.92)	0.48 (0.33, 0.60)
NNDS Score - BPRS Score	0.80 (0.77, 0.84)	0.58 (0.45, 0.69)	0.70 (0.61, 0.78)	0.06 (-0.12, 0.23)

NNDS - NIMHANS-NMHS2 Disability Scale, BPRS=Brief psychiatric rating scale, WHODAS=WHO-Disability Assessment Schedule

K-means cluster analysis, performed on the combined WHO-DAS 2.0 and NNDS scores, categorized the participants into four distinct clusters based on severity of disability. As detailed in Table 5 and Figure 1, the mean NNDS scores for these clusters ranged from 13.81 to 45.55, with an increasing severity of disability from cluster 1 to cluster 4. The distribution of participant groups across these clusters showed a clear progression, with the community sample dominating Cluster 1 (no disability) and patients dominating Clusters 2, 3, and 4. Specifically, outpatients were spread across Clusters 1–3 (notably in Cluster 1 and 2- no/mild disability), while inpatients were concentrated in Clusters 3–4 (moderate/severe disability).

Table 5.

K-Means of disability scores across clusters and the distribution of each cluster participants among the three settings

NNDS score	Cluster 1 (n=216)	Cluster 2 (n=89)	Cluster 3 (n=91)	Cluster 4 (n=20)
Mean±SD; Median (Q1, Q3); Min-max	13.81±1.53 13.00 (13.00, 14.00) 13–20	23.89±4.07 23.00 (21.00, 26.00) 16–37	34.66±3.56 35.00 (32.00, 38.00) 26–42	45.55±5.42 46.00 (41.25, 50.00) 34–52
Group
Community (n=133)	130	2	1	0
Outpatients (n=152)	68	47	28	9
Inpatients (n=131)	18	40	62	11

Min-Max: Minimum – Maximum; NNDS score: NIMHANS-NMHS-2 disability scale scores

Figure 1.

Distribution of scores

ROC analyses of NNDS score based on the four clusters depicted in Figure 2, identified optimal cutoff points to classify the four distinct clusters. The Area Under Curve (AUC) values show excellent discrimination between clusters. The defined cutoffs (18, 31, 40) correspond with transitions from Cluster 1 to 2, Cluster 2 to 3, and Cluster 3 to 4. The optimal cut-offs thus determined for the NNDS were: No disability: 13–18; Mild disability: 19–31; Moderate disability: 32–40; Severe Disability: >40.

Figure 2.

ROC analyses of NNDS score based on the four clusters. AUC - Area under the curve - Numbers represent the AUC with 95% confidence intervals. NNDS = NIMHANS-NMHS-2 Disability Scale, ROC = Receiver operating characteristic

DISCUSSION

The NIMHANS-NMHS-2 disability scale showed strong psychometric properties in adults, characterized by excellent levels of internal consistency and a stable 5-factors structure. It showed very strong correlation with the gold standard for measure of disability (WHODAS) (rho = 0.93) and measure of psychopathology (BPRS) (rho = 0.80). The NNDS had a near-perfect joint inter-rater reliability. The study sample had participants with the full spectrum of severity, ranging from those in the community sample with no diagnosable mental health disorders to out-patients with mild-moderate symptoms and in-patients with severe symptoms. Additionally, the patient sample consisted of six major psychiatric diagnostic categories in treatment settings. The rigor applied during the development of the NNDS, including a thorough focused desk review, cognitive interviews from stakeholders and content validation by experienced consultants has significantly contributed to the high levels of agreements observed.

The findings from this study are consistent with previous studies examining the psychometric properties of two widely used disability scales in our country—the WHO-DAS 2.0 and IDEAS. The 36 and 12 item-WHODAS is the most extensively studied; pooled findings has demonstrated high internal consistency (α: 0.86), a stable factor structure; high test-retest reliability (ICC: 0.98) and good concurrent validity.[9] The IDEAS scale has also demonstrated good internal consistency (α = 0.78), and criterion validity ranging from 0.7 to 0.8 across multiple studies comparing it with other disability scales.[19,20,21,22]

Although being intended to evaluate six functional areas, exploratory component analysis showed that the NNDS has a stable five-factor structure. This consolidation occurred due to significant overlap between the participation and interpersonal functioning domains [Table 3]. This convergence is clinically meaningful, as the capacity for effective communication and relationship maintenance is intrinsically linked to an individual’s ability to engage in community activities and social decision making. A systematic review of the 12-item WHO-DAS 2.0[23] has also identified a multidimensional structure with up to five factors.

Establishing reliable cut-off scores for disability severity has significant clinical utility. While our results require replication, it was possible for our study to successfully categorize participants into four distinct groups. The categorization aligned with expected clinical trajectories, showing a linear progression in disability scores from community, outpatient, and inpatient samples. The analysis also showed excellent cluster discrimination, characterized by well-defined transitions and clear participant categorization. However, it has not been possible to calculate the cut-off scores for profound disability as very few people in our sample had profound disability. This aligns with previous research suggesting that profound disability is rare in mental illness, typically occurring only in the context of co-occurring developmental disorders.[21]

According to the World Health Report in 2001, nearly one-third of the global disability attributed to mental disorders, with major depression, alcohol dependence, schizophrenia, bipolar affective disorder and obsessive-compulsive disorder among the top 10 causes of disability.[24] The 2019 Global Burden of Disease Report confirmed these findings, noting no evidence of a reduction in that burden over the last 3 decades.[25] Therefore, the assessment of disability represents a primary objective of the NMHS-2. Quantifying disability is also vital for effective service planning and the optimal allocation of constrained resources. The NNDS, developed and validated specifically for mental health conditions and incorporating cultural nuances of our country, might guarantee a more rigorous and contextually appropriate evaluation for the NMHS-2.

The assessment of disability in the NMHS-2 is envisaged to have several other unique contributions. The survey will assess all disorders with public health importance, and the resulting data will tell us the severity and patterns of disability associated with these disorders. While disability across various psychiatric disorders has been well-documented in clinical samples, this survey will be the first study to examine it in a nationwide community sample. The NMHS-2 assesses all participants, in contrast to most mental health epidemiological surveys that limit disability assessment to identified cases. This allows establishing disability norms for those healthy, those with sub-threshold symptoms, and those with diagnosed mental health disorders.

A major limitation of this study is that findings were derived from a single center, which necessitates replication across multiple sites. Individuals with dementia, autism and disorders of intellectual development were excluded from the sample, as this would have necessitated primarily a care-giver proxy assessment. Test-retest reliability could not be established as most in-patients improved clinically within one month. Consequently, their perception of functioning changed rapidly to allow for meaningful assessment of the scale’s temporal stability. Although not to be considered a study limitation, the community sample showed no significant correlation with the measure of psychopathology (BPRS), despite very strong overall correlations. This was primarily owing to restricted range of scores among the control sample, which provided insufficient variability required for a robust statistical analysis.[26]

While this study validated the interviewer-administered adult version, future studies should examine its validity for self-administration, its psychometric properties among adolescents and translated versions. Additionally, as the limited sample size precluded an analysis of individual diagnostic categories, this remains as an objective of future research.

The NMHS-2 planning process prioritized a dual approach: ensuring a comprehensive multi-domain assessment without burdening participants by optimizing administration time. The 13-item NNDS hence underwent further evaluation to determine if a condensed version could be developed while maintaining its full psychometric integrity. This resulted in the 8-item NNDS-SV, which is currently being utilized in the survey. The manuscript regarding its development and validation is being separately published.

In summary, the NIMHANS-NMHS-2 disability scale have demonstrated robust psychometric properties for measuring psychiatric disability. With the instrument being open-access, the scale offers broad applicability for both clinical practice and community-based research nationwide.

Ethics

The study received ethical clearance from the Institute Ethics Committee (Behavioral Science Division), NIMHANS (No. NIMHANS/EC (BEH.SC.DIV.) MEETING/2023 dated 19.4.2023. Participants provided written informed consent prior to assessment.

Conflicts of interest

There are no conflicts of interest.

Funding Statement

Funded by the Department of Health and Family Welfare, Government of India through the National Mental Health Survey -2.