Validation of Bengali version of Edmonton Symptom Assessment Scale-Revised (ESAS-r Bengali): A multidimensional symptom assessment tool for patients with advanced incurable diseases receiving palliative care

Abstract

Objective:

Routine symptom assessment represents the cornerstone of symptom management of patients with advanced incurable diseases in palliative care. At present, there is no validated tool to assess symptoms among the Bengali-speaking population with incurable diseases. The aim of the study is to translate, culturally adapt, and validate the Edmonton Symptom Assessment Scale (Revised) into Bengali language.

Methods:

The study was conducted in two phases. Forward and backward translations of the English version of the Edmonton Symptom Assessment Scale (Revised) into Bengali were conducted by four independent translators. After obtaining reviews from an expert committee, pre-testing and cognitive debriefing the Bengali version of the tool was finalized. The final validation was conducted among 110 patients admitted to the Palliative Medicine Department of Bangabandhu Sheikh Mujib Medical University, Dhaka, Bangladesh. Validity (content, face, and construct validity) and reliability (internal consistency) were assessed in the final validation phase.

Result:

All participants responded to all items. Seventy percent of the participants completely understood all questions but 30% had difficulty with three questions. The expert committee expressed their satisfaction regarding the face and content validity of the questionnaire. The Bengali version also had high reliability (α = 0.862). Principal component analysis with the distribution of varimax rotation of Edmonton Symptom Assessment Scale (Revised) Bengali ranged from 0.41 to 0.83.

Conclusion:

After the translation and cross-cultural adaptation, the Bengali version of the Edmonton Symptom Assessment Scale (Revised) achieved good levels of validity and reliability. It can be used as a symptom assessment tool for Bengali speaking population receiving palliative care.

Introduction

Patients with advanced incurable diseases experience significant symptom burden from the time of diagnosis onwards, and often increasing in intensity over time. These symptoms are often multidimensional in nature and can negatively impact patients’ quality of life and performance status. ¹ The goal of palliative care is to improve the quality of life by controlling the symptoms and integrating the psychosocial and spiritual aspects of patient care. ² Common symptoms like fatigue, pain, shortness of breath (SOB), drowsiness, appetite loss, depression, and anxiety were the most frequent symptoms experienced by >50% of the patients during the last 1–2 weeks of life. ³

One of the most critical aspects of symptom management in palliative care is routine symptom assessment and re-assessment—which allows symptoms to be recognized, diagnosed, treated, and monitored over time. Control of symptoms becomes more efficient if the assessment is done based on a systematic and standardized tool. ⁴

There are multiple tools available for symptom assessment in palliative care patients. Among them, the Edmonton Symptom Assessment Scale (ESAS) is one of the first quantitative symptom assessment tools that allows for simple and rapid documentation of multiple patient-reported symptoms at the same time. ⁴ First developed in Canada, this tool has evolved into an important symptom assessment instrument in both clinical practice and research. The current revised version Edmonton Symptom Assessment Scale-Revised (ESAS-r), assesses the severity of nine common symptoms: pain, tiredness, drowsiness, nausea, SOB, appetite, depression, anxiety, and well-being. ⁵ It can also be used to assess the nature of symptom trajectory, symptom clusters, and symptom modulators. ⁴ This tool is commonly used for symptom screening and longitudinal monitoring of the symptoms not only in patients of palliative care but also in oncology, nephrology, and other disciplines in both inpatient and outpatient settings. ⁶

A large number of people with life-limiting illnesses are seeking palliative care in Bangladesh. There is no validated assessment tool for this Bengali-speaking population to assess their symptom burden. This study aims to translate the English version of the ESAS-r in Bengali followed by standard validation.

Methods

Edmonton Symptom Assessment System—Revised

The ESAS-r assesses nine common symptoms: SOB, pain, fatigue, nausea, depression, anxiety, drowsiness, lack of appetite, and overall well-being, with the option of adding a tenth patient-specific symptom. There are visual analog scales for each symptom, ranging from 0 (no symptom) to 10 (worst possible symptom). Higher scores represent higher symptom intensity. The time frame for symptom ratings is specified as “now” referring to at the time of application of the instrument.

Translation procedure

The translation process from English to Bengali was performed according to the Beaton et al. ⁷ guidelines. The forward translation was carried out by two independent Bengali-speaking translators one of whom was aware of the concept of the scale (T1), another was not (T2). One of the forward translators was from medical background (T1), and another one was an expert in English literature from non-medical background (T2). These two versions were then synthesized by another third unbiased person into a combined one (Ts). Only a few changes in the Ts version had required since it covered almost all local customs, habits, and usage of words. The Ts version of the questionnaire was back-translated into English by two independent translators (BT1 and BT2) with good command in the English language. Both of them were experts in English literature and from non-medical backgrounds. These two back translators were kept uninformed about the concept of the tool to avoid information bias and to elicit unexpected meanings of the items in the translated questionnaire. The back-translation process magnified unclear wording in the translations. Then an expert committee was formed with three Palliative Medicine experts from the Palliative Medicine Department and a psychiatrist from the Psychiatry Department of Bangabandhu Sheikh Mujib Medical University (BSMMU). All of them were trained in the English version of ESAS. The expert committee and all translators reviewed and compared all the translations and the original ESAS in August 2022. They verified the semantic, idiomatic, experiential, and conceptual equivalence between the English and Bengali versions. After reaching a consensus among all the members of the committee, the final Bengali version was made for pre-testing (Table 1).

Table 1.

Translation process of ESAS-r Bengali.

Stage I: Translation	Two translations (T1 and T2)	Written report for each version (T1 and T2)
	Into target language
	Informed and uninformed translators
Stage II: Synthesis	Synthesize T1 and T2 into Ts	Written report
	Resolve any discrepancies with the translator’s reports
Stage III: Back translation	Two re-translated versions from Bengali to English translations by proficient English teachers	Written report for each version (B1 and B2)
	Work from Ts version
	Create two back translations (B1 and B2)
Stage IV: Expert committee review	Review all reports	Written report
	Methodologists, developers, language professionals, translators
	Reach a consensus on discrepancies
	Produce pre-final version
Stage V: Pilot-testing	n = 30 (children), n = 10 (patients)	Written report
	Complete questionnaire
	Probe to get an understanding of the item

Pilot-testing

The pre-testing for the final Bengali version of the ESAS-r was administrated to 30 children of Grade VI (roughly 12 years of age) of Dhaka Residential Model College. Each question was presented to the subjects and asked to describe what they understood by the question and how they would answer if the condition were present in them. The pre-tested questionnaire was then applied among 10 Bengali-speaking patients admitted to the Palliative Medicine Department of BSMMU in September 2022 for cognitive debriefing. Their data was only used for this procedure, not for the main study. The patients were asked about any difficult, confusing, or upsetting words during the administration of the questionnaire. Alternative words and brief explanations were offered in the case of difficult or confusing words/phrases. After reviewing the revised questionnaire by the expert committee the final version was approved.

Setting and participants for the final validation phase

We enrolled 110 patients for the final validation phase. The sample size was calculated based on item and sample ratio (1:10). As the total item number is 10 total of 100 samples were considered. For consideration of a 10% drop out a total of 110 respondents were approached. So, the final sample size was 110. Patients with advanced incurable diseases admitted at the Palliative Medicine Department of BSMMU, aged ⩾18 years, who could read and understand Bengali, were included in the study. Patients having any cognitive impairment or psychotic disorders were excluded from the study.

Data collection

Data collection was carried out from October 2022 to December 2022. The samples were obtained by a consecutive sampling method. Eligible patients were requested to fill up the questionnaire by themselves. If any confusion rose regarding any of the items, the principal investigator gave them brief explanations for the items as needed. Each questionnaire completion lasted 5–10 min.

Data analysis

Face validity was assessed during the standard translation process. Content validity was also assessed during standard translation, back-translation, expert committee review, and literature reviews. Construct validity was assessed by factor analysis with principal component analysis with varimax rotation. Items loading <0.30 were considered to be discarded. Reliability was assessed by Cronbach’s alpha coefficient and the standard was taken as ⩾0.70. Data analysis was performed by Statistical Package for the Social Science (SPSS), version 26, manufactured by IBM Corporation.

Ethical considerations

Ethical approval for both the research and consent procedure (approval no: BSMMU/2022/5550, date: June 04, 2022) was obtained from the Institutional Review Board, BSMMU. Written informed consent was taken from all eligible patients. As they were terminally ill patients, their health conditions were considered during data collection.

Result

Validity analysis

Face validity: The process of face validity was also built into the translation and adaptation method. The members of the expert committee and the naive respondents (uninformed translators) both agreed that the purpose of the test was clear and it should assess symptoms of the patients with incurable illnesses. During the pre-testing phase, out of 30 children, 85% were able to understand all the questions and the remaining 15% needed some explanation. During cognitive debriefing, out of 10 patients, 70% completely understood 7 questions (70%) and 30% understood all questions. All components of the Bengali version were relevant to what was being measured. Only items “lack of appetite,” “depression,” and “well-being” needed brief explanations. The measure and the measurement method were useful and appropriate in capturing and measuring the variables. The expert committee considered the final version of the instrument as an accurate tool for evaluating the symptoms among the Bengali-speaking population. No separate validation group was needed to test out the revised version.

Content validity: Content validity of the final version was assessed by three experts: two Palliative Medicine experts and one psychiatrist. Item-level content validity index (I-CVI) was found to be 1 (table-4) for each item and scale-level content validity index (S-CVI) was therefore 1 by the averaging calculation method (Table 2).

Table 2.

Item descriptive of content validity of the adapted Bengali version of the ESAS-r.

Item	Rating by experts			I-CVI	S-CVI
	Expert 1	Expert 2	Expert 3
1	4	4	4	1	0.93
2	4	4	4	1
3	4	4	3	0.66
4	4	4	4	1
5	3	4	4	0.66
6	4	4	4	1
7	4	4	4	1
8	4	4	4	1
9	4	4	4	1
10	4	4	4	1

Construct validity: A nearly equal number of participants from both sexes, the majority aged between 41 and 50 years, took part in the final validation process. Most of them (80%) had education up to secondary to higher secondary level (Table 3). Construct validity of the final version was assessed by exploratory factor analysis with principal component with varimax rotation. The observed Kaiser-Meyer-Olkin (KMO) value of 0.72 indicating the present data was adequate for the factor analysis. Bartlett’s test of sphericity (χ² = 244.42, df = 45, p < 0.001) revealed that the present sample was suitable for factor analysis. ⁸ The distribution of varimax rotation of ESAS-r Bengali ranged from 0.41 to 0.83. The highest score was 0.83 for the item 5 (lack of appetite). The lowest score was 0.41 for item 3 (drowsiness) (Table 4). There was only one component comprising the 10 items was extracted.

Table 3.

Demographic variables of the participants (n = 110).

Variable	Participants
	Frequency	Percentage
Age groups
18–30 years	21	19.09
31–40 years	32	29.09
41–50 years	39	35.5
51–60 years	18	16.36
Education status
Primary level	11	10
Secondary level	48	43.64
Higher secondary level	40	36.36
Graduation	11	10.0
Sex
Male	51	46.36
Female	59	53.63

Table 4.

Principal component analysis with the distribution of varimax rotation of ESAS-r Bengali.

Component matrix
Items	Component 1
Q1	0.619
Q2	0.801
Q3	0.415
Q4	0.730
Q5	0.839
Q6	0.765
Q7	0.666
Q8	0.604
Q9	0.810
Q10	0.621
Extraction method: principal component analysis
1 component extracted

Reliability analysis

The Cronbach’s alpha (α) value ⩾0.70 was considered as adequate and ⩾0.80 was considered as optimal. In our study, the Cronbach’s alpha (α) value for ESAS-r Bengali was 0.862 (Table 5).

Table 5.

Internal consistency of the ESAS-r Bengali.

Reliability statistics	Value
Cronbach’s alpha	0.862
Cronbach’s alpha based on standardized items	0.811
Number of items	10

Discussion

Translation and standardization of a scale are crucial components of cultural adaptation of a standard scale or instrument. Finding appropriate words in translated language sometimes become a great challenge for translators. In this study, we have translated and validated the ESAS-r scale which is one of the most important tools for symptom assessment in the field of palliative care.

The reliability of ESAS has been assessed by multiple research groups. Bruera et al. ⁹ found that ESAS had good test-retest reliability among 34 hospitalized patients, and correlated with Support Team Assessment Schedule. Philip et al. ¹⁰ assessed the validity of a slightly modified version of ESAS assessing symptoms in which “activity” was replaced with “weakness” in 80 patients with cancer from Australia. ESAS had a satisfactory to good correlation with Brief Pain Inventory and Rotterdam Symptom Checklist, with weighted kappas between 0.46 and 0.61. ¹⁰ We have followed the standard validation method of the original English version during translating and culturally adapting the tool into Bengali language.

The internal consistency of the Bengali version was 0.862, which denotes high reliability. It is as reliable as other language versions such as Icelandic (α = 0.85), Chinese (α = 0.79), and Thai (α = 0.75).^6,11,12

The face validity of this tool has also been found to be highly consistent with the original version. The expert committee found the version accurate for the symptom assessment of the Bengali-speaking population. During cognitive debriefing, three items (lack of appetite, depression, and well-being) needed brief explanations.

Item number 5 is “No Lack of Appetite.” There are multiple synonyms and expressions for the term “No Lack of Appetite.” Usually in Bengali when two negative words are used at a time, it becomes difficult to understand for the patient. Also, patients with lower educational status needed either further explanation or elaboration of the meaning of this particular phrase. Other studies showed that patients who evaluated the ESAS stated that the term “No Lack” was a double negative and therefore led to confusion. ¹³

Item number 7 is “No Depression.” In Bengali, the word depression is used as a synonym for sadness. So we needed to elaborate on this symptom using phrases like “no pleasure in life” or “not finding any desire to do daily activity” to differentiate depression from sadness. The word “depression” is used by the general population to describe “sadness” and a symptom of mental illness interchangeably. ¹⁴ This word is needed to be reevaluated because many patients had mistaken “depression” for “sadness.” This confusion needs to be clarified during the intervention, as there are some differences between the treatment plan for the patients who “feel a lack of joy” and who are suffering from “clinical depression.” ¹⁵

Item number 9 is “Best Well-Being.” The term “well-being” seems vague and non-specific to many of our participants. A further explanation was needed to clear their confusion. So the expert committee decided to use a conceptual synonym for the meaning of well-being. Our situation is similar to another study, where the meaning of the word “well-being” was not understood by the patients and hindered the overall rating of the instrument. ¹³

About the term “Drowsiness = Feeling Sleepy,” the Bengali dictionary states that the meaning of “sleepy” is “the act of sleeping.” One study changed the term “drowsiness” to “sleep.”^14,16 During the synthesis phase, all members of the expert committee approved the translation “feeling sleepy,” because it does not alter the meaning of the words, implying that if the person who completes the instrument does not understand the meaning of the word drowsiness (sleepiness) and consults the definition, this person will not understand the expression of drowsiness (feeling sleepy) either. The wording did not need to be changed during the validation phase.

One comment, which was also made in the group of patients, was related to the symptom “nausea.” The inclusion of this expression was suggested to facilitate the understanding of patients who might be embarrassed to ask for the definition of this word. Another suggestion, mentioned previously, was related to the standardization of the expression “lack = without” in the left column of the instrument. This alteration was suggested to improve the perception of the instrument. None of the studies on ESAS mention this specific detail. Consequently, this change was not thought necessary, considering that none of the patients expressed confusion when they graded the items.

Exploratory factor analysis confirmed the existence of a single component of this scale which was consistent with the original English version. Except for item number 3 “drowsiness,” the factor loads of the rest of the items were more than 0.60 which indicates a very good correlation. All findings are consistent with the results of previous studies. ¹¹

During data collection, we faced difficulty with the timing of the symptoms, whether it was “before or after getting palliative treatment,” “at the time of admission,” or “at the time of data collection.” We had to clarify that the word “now” refers only to the symptoms present “at the time of data collection.” This confusion was raised during the administration of the Thai version as well. ⁶

One limitation of this study is that we couldn’t perform test-retest reliability as many of our patients were nearly at the end of life, so repeating this questionnaire was difficult for them. Another limitation is that we couldn’t assess the convergent validation of this tool, as there is no other tool in Bengali for symptom assessment for palliative patients.

Conclusion

The Bengali version of ESAS-r was found to be valid and reliable as well as culturally appreciable for the Bengali- speaking population. So, it can be applied in clinical and research settings for the assessment of symptoms and seeing treatment responses of patients suffering from advanced incurable illnesses receiving palliative care in Bengali-speaking countries.