Development of a distress inventory for cancer: preliminary results.B Thomas, VN Mohan, I Thomas, M Pandey
Department of Futures Studies, University of Kerala, and Division of Surgical Oncology, Regional Cancer Centre, Trivandrum, India. , India
Correspondence Address: Source of Support: None, Conflict of Interest: None PMID: 12082320
Source of Support: None, Conflict of Interest: None
CONTEXT: Advances in cancer treatment have led to cure and prolongation of patients' lives; however associated psychosocial problems, including distress, can detrimentally affect patients' compliance with treatment and ultimately, their outcome. Symptom distress has been well addressed in many studies; however, psychological distress has only been quantified by using depression or anxiety scales/checklists or quality of life scales containing a distress sub scale/component or by the use of scales that are not psychological distress-specific. AIMS: The present study is an attempt to construct a psychological distress inventory for specific use with cancer patients. SETTINGS AND DESIGN: The standardisation sample consisted of 63 randomly selected patients with head and neck cancer who had undergone/ were undergoing curative treatment at the Regional Cancer Centre, Trivandrum. PATIENTS AND METHODS: The Distress Inventory for Cancer contained 57 positively and negatively toned items. An item analysis was conducted, followed by a factor analysis, thereby identifying the domains influencing distress. RESULTS: The final questionnaire contained 26 items subdivided into four domains viz. the personal, spiritual, physical, and the family domains, with each domain providing a sub score. The reliability coefficient (Cronbach's alpha) of the scale was found to be 0.85. CONCLUSIONS: These are the preliminary results of an ongoing study on global distress and tool development process. Reported here is the first step towards development of such tool.
Keywords: Adolescent, Adult, Aged, Aged, 80 and over, Factor Analysis, Statistical, Female, Head and Neck Neoplasms, psychology,Human, Male, Middle Age, Psychiatric Status Rating Scales, Quality of Life, Questionnaires, Self Assessment (Psychology), Stress, Psychological, diagnosis,
Recent advances in diagnosis and treatment of cancer has led to a significant increase in cancer survival. However, in pursuit of longevity the ‘quality’ of survival is often overlooked. With a diagnosis of cancer, all patients irrespective of age, and sex are distressed. In specific reference to cancer, distress is defined as “an unpleasant emotional experience of a psychological (cognitive, behavioural, emotional), social, and/or spiritual nature that interferes with the ability to cope effectively with cancer and its treatment. Distress extends along a continuum, ranging from common normal feelings of vulnerability, sadness and fears, to problems that can become disabling, such as depression, anxiety, panic, social isolation, and spiritual crisis.”
Screening for distress has been documented as a dire need for cancer patients.,,, Besides, distress is not easy to identify, as the patients may be reluctant to spontaneously disclose their distress to oncologists, or the oncology staff may be too busy to notice sub-clinical distress. Therefore, a viable solution is to develop paper and pencil screening techniques, as they are inexpensive and convenient methods to identify distress, and help to intervene in select cases.
Review of literature shows that as many as 45 tools/instruments have been used to measure psychological distress, commonest being ‘Brief Symptom Inventory’ (BSI). The BSI was designed to reflect the psychological symptom patterns of psychiatric and medical patients as well as community (non-patient) samples. ‘Hospital Anxiety and Depression Scale’, a scale measuring anxiety and depression is validated in identifying cases against the criterion of a psychiatric assessment. The ‘General Health Questionnaire’, was developed for detecting independently verifiable psychiatric morbidity (generally anxiety and depression). However, none of these scales are able to identify patients who are highly distressed but do not have clinical symptoms of anxiety and depression.
The ‘Psychological Distress Inventory’ developed by Morasso et al and used in three other studies,,, is a validated self-administered, cancer-specific Italian distress questionnaire. The present study reports the results of the first phase of the tool development process for measuring global distress in cancer patients.
A review of literature, patient interviews, and discussions with the experts in the field led to the identification of six dimensions that could possibly influence the distress response. These dimensions were: psychological, social support, general health and treatment, family relationship, self-confidence, and spirituality. Using these dimensions as content areas, items were framed to index the psychological distress in cancer patients. Selection of these six dimensions ensured content validity of the tool. However, these were not used as sub-scales in the inventory. The dimensions served only as a guideline in the development of the items.
The responses to the items were measured on a five point Likert scale. There were both positively and negatively toned items so as to control for the effects of acquiescent responding. Care was taken to ensure that there were no ambiguously stated, or loaded items.
The draft scale labelled Distress Inventory - Cancer (DI-C) contained a set of instructions for the respondent and a total of 57 items. The DI-C was bilingual (English followed by the local dialect - Malayalam). Of the 57 randomly arranged items, 25 were positively toned and the remaining negative. The instructions to the respondent contained the necessary information regarding the response options, and how to indicate one’s response. The questionnaire took approximately 15-20 minutes to complete.
A total of 68 patients newly diagnosed with head and neck cancer, awaiting curative treatment in form of surgery or radiotherapy were interviewed. Five incomplete questionnaires were rejected and hence the final standardisation sample consisted of 63 patients (age ranged from 18-88 years, and 14.3% of the sample were females). The demographic profile of the standardisation sample is detailed in [Table - 1].
The instructions were read out to the patients willing to participate in the study. The patients followed these in his/her test booklet. After clearing doubts, if any, the patients were asked to fill in the questionnaire. Patients who could manage on their own used the tool as a self-report scale, while for those who could not read or had trouble reading, the items were read out and the response recorded by the investigator. All the interviews were carried out by a single investigator (BCT).
A weight of 5, 4, 3, 2, or 1 was given to a positive item. The scores were reversed if the item was negatively toned. The response sheet was not scored if there were more than one response category chosen for a particular item or if there were three or more omitted items. If there were only one or two omitted items, a score of 3 was given for each. The maximum score obtainable was 285 and the minimum, 57.
The item selection procedure was divided into two phases, viz., i) Item analysis, and ii) factor analysis. The item analysis was carried out by determining the impact of each item on the overall Cronbach’s alpha of the tool and was based on the exclusion of the particular item. The factor analysis was carried out by principal component method of extraction and varimax rotation. Reliability of the final inventory was estimated using Cronbach’s alpha.
The mean score obtained was 181 and it ranged from 119 to 248. The item analysis involving Cronbach’s alpha for each item indicated that all the 57 items had high individual alpha scores. Therefore, all the 57 items were retained for the Factor Analysis.
The first stage of factor analysis yielded 16 factors. The factor analysis was repeated after restricting the extraction of factors to those with eigen value >2. From the six factors thus obtained, items with factor loading <0.55 were deleted resulting in the removal of 22 items. The remaining items were again subjected to repeat factor analysis with the eigen value of $2, and factor loading $0.55. This resulted in the extraction of four factors with a total of 26 items [Table - 2] - Factor I and III having 7 items, and Factor II and IV with 6 items each. The four factors obtained were studied by the experts and based on their consensus these were appropriately named. Factor I was named as ‘Psychological domain’, Factor II as the ‘Spiritual domain’, Factor III as the ‘Physical domain’, and Factor IV as the ‘Family domain’.
Cronbach’s alpha for the DI-C was found to be 0.85. The alpha scores for the four factors namely; Psychological
domain, Spiritual domain, Physical domain, and the Family domain were 0.85, 0.52, 0.79, and 0.72 respectively.
As the DI-C contains items that appear to measure the variables in question, the inventory can be said to have face validity. The items for the DI-C were based on the six content areas identified through literature, patient interviews, and discussion with experts in the field; hence, the inventory has content validity. Factor analysis of the items in the inventory has led to the identification of four principal components that were theoretically and psychologically meaningful. Hence, the inventory can be said to have factorial validity as well.
The final tool had the 26 items rearranging themselves into the four factors [Table - 2]. The Personal domain alone explained 28% of the total variance, with four factors together accounting for 59% of the variance. The domains themselves gave sub-scale scores and when combined would give the global distress score. The minimum and maximum scores obtainable is 26 and 130 respectively. In the study sample, the scores ranged from 50 to 111 with the mean score being 80.5. The minimum and maximum obtainable scores for the psychological and family (first and fourth) subscales is 7 and 35 respectively while for the spiritual and physical (second and third) subscales, it is 6 and 30. The mean scores (range) obtained in the present study for various domains were 26.8 (16–35), 18.2 (9–26), 24.6 (6–28), and 20.9 (10–31) respectively for psychological, spiritual, physical, and family domains.
Normal emotions like fear, worry, and sadness occurs in every person, and exacerbate with the occurrence of a diagnosis of cancer. Clinical entities like anxiety and depression do not develop overnight; instead, they are a cumulative outcome along a continuum that extends beyond normal psychological and emotional reactions. Distress is therefore beyond normal feelings of sadness and fear, and could be termed ‘subsyndromal’ as it does not quite reach the clinical entity like anxiety or depression. Anxiety can be overcome by use of anxiolytics, however distress needs psychological intervention. The need for supportive counselling and psychological intervention and its positive impact on stress has been highlighted in a number of studies.,,,,,,,,,,, In the developing countries, it is difficult to provide these services to every patient who is registered at the cancer treatment centre, as the number of patients is very high and resources are limited. Therefore, the development of a screening mechanism seems appropriate and may help in identifying those who specifically warrant an intervention. Besides this, distress screening may assist health professionals to provide patient-specific-intervention processes if the distress level and cause could be identified.
Therefore as part of the first phase towards developing a distress-screening tool to estimate the thresholds along the
distress continuum, a tool to identify the patient’s psychological distress level was developed. The tool developed is
bilingual i.e. in English and Malayalam, however, only the Malayalam version has been validated.
The Distress Inventory – Cancer reported here measures the ‘Global Distress’ of cancer patients in terms of four
domains, viz. Psychological domain, Spiritual domain, Physical domain, and the Family domain. It was noted that though the social and self-confidence dimensions identified by the subject experts were removed by the factor analysis, items from these had rearranged into the Family, Spiritual, and Psychological domains. This could be due to the role of family as primary source of social support,,, and that the patient’s spirituality has a major impact on his/ her confidence and coping with cancer.,,,
The present distress inventory has been found to be suitably reliable and valid and therefore appropriate for the measurement of the distress levels of the cancer patient. Further studies are on to develop the percentiles and for further validation of final tool.
In the present form, DI-C is well suited for research studies dealing with distress and its correlates. However, its use in the clinical setting as a screening tool requires well-established normative scores. Though the psychometric properties of the test (content validity, factorial validity, and internal consistency) have been well established for the tool, there is scope for further development in the precision of measurement. The present study is limited mainly due to the small and homogenous sample size. Additional studies will also be required to establish the construct validity and the test-retest reliability. The English language tool needs to be tested in an English speaking population and across a wide social and cultural set-up to establish its cross-cultural linguistic utility.
The authors would like to acknowledge Dr. Iqbal M. Ahmad, Director, Malabar Cancer Centre, Tellichery, Dr. Paul Sebastian, Associate Professor and Head, Surgical Oncology, Dr. K. Ramadas, Associate Professor, Radiation Oncology, and Prof. M. Krishnan Nair, Director, Regional Cancer Centre, Trivandrum for their help in carrying out this study.
[Table - 1], [Table - 2]