|Year : 2015 | Volume
| Issue : 3 | Page : 234-238
Psychometric properties of the Malayalam version of ECOHIS
Sangeetha Govinda Bhat1, Rema Sivaram2
1 Assistant Professor, Department of Pedodontics and Preventive Dentistry, Amrita School of Dentistry, Ponekkara, Cochin, Kerala, India
2 Honorary Faculty, Clinical Epidemiology Resource and Training Centre, Government Medical College, Thiruvananthapuram, Kerala, India
|Date of Web Publication||9-Jul-2015|
Dr. Sangeetha Govinda Bhat
Assistant Professor, Department of Pediatric Dentistry, Amrita School of Dentistry, Ponekkara P.O. Cochin - 682 017, Kerala
Source of Support: None, Conflict of Interest: None
| Abstract|| |
Background: The Early Childhood Oral Health Impact Scale (ECOHIS) has been developed in English for determining oral health-related quality of life (OHRQoL) in the preschool children. It has been translated and validated in different languages to suit different cultures. The ECOHIS is, till date, the only tool available for research in this field on preschool age children. A similar version of this tool is not available for use in the local language Malayalam. Aim: This study aimed to develop and validate a Malayalam version of the ECOHIS (M-ECOHIS). Design: The study was conducted with a cross-sectional design. The ECOHIS was translated into Malayalam by forward-backward translation and tested for face and content validity. The parents of 300 children were administered the M-ECOHIS and an additional global oral health (GOH) question. The children were examined for the presence of early childhood caries (ECC) which was recorded using the defs index. The internal consistency reliability, test-retest reliability, interobserver reliability, and convergent and discriminant validity were assessed. Results: The tool possessed good internal consistency (Cronbach's alpha = 0.879; item total correlation 0.2832-0.7617); the test-retest reliability and interobserver reliability assessed using ICC (ICC = 0.9457 and 0.9460, respectively) was acceptable. The Spearman's correlation coefficient of the ECOHIS and the GOH scores, r = 0.725; P = 0.01 supported the convergent validity. The mean ECOHIS scores of children having ECC with mean ECOHIS scores of children without ECC were compared using unpaired t-test and found to be statistically significant supporting the discriminant validity of the scale. Conclusion: The M-ECOHIS can be used to assess the OHRQoL of preschool children in the Malayalam speaking community.
Keywords: Ecohis, ECC, tool, validation
|How to cite this article:|
Bhat SG, Sivaram R. Psychometric properties of the Malayalam version of ECOHIS. J Indian Soc Pedod Prev Dent 2015;33:234-8
| Background|| |
The effects of early childhood caries (ECC) in young children extend beyond the mouth. Tooth loss is sometimes inevitable, and it can cause not only orthodontic and esthetic problems, but more importantly, difficulties in pronunciation. Esthetic problems and pronunciation difficulties may result in psychological and relationship problems. In addition, children with ECC usually weigh less and are shorter than average. , Their growth is affected because they have difficulty in sleeping and eating as a result of the infection and pain, and their quality of life is greatly diminished. 
In the preamble of its constitution, World Health Organization (WHO) states that 'health is a state of complete physical, mental and social well-being and not merely the absence of disease and infirmity'  In dentistry, this new perspective on health suggests that the ultimate goal of dental care is not merely the absence of caries or periodontal disease or oral cancer but also the mental and social well-being of the patient. The concept of oral health-related quality of life (OHRQoL) captures the aim of this new perspective.
Multiple items questionnaires are the most widely used method to assess OHRQoL. One such instrument is the early childhood oral health impact scale (ECOHIS). It is a short instrument to be completed by the parent or the primary caregiver. It is structurally composed of 13 items distributed between two sections: The Child Impact Section (CIS) and Family Impact Section (FIS). The CIS has four subscales: Child symptom, child function, child psychology, and child self-image and social interaction. The FIS has two subscales: Parental distress and family function. The scale has five rating response options to record how often an event has occurred in the life of the child: 0 = never; 1 = hardly ever; 2 = occasionally; 3 = often; 4 = very often; and 5 = don't know. ECOHIS scores are calculated as a simple sum of the response codes for the CIS and FIS. CIS and FIS scores range from 0 to 36 and 0 to 16, respectively. Higher scores indicate a greater oral health impact and poorer OHRQoL. The instrument is intended for use in the epidemiological surveys to assess the burden of dental disease and its treatment among preschool aged children at a population level. Developed by Pahel et al.,(2007)  in the USA, the tool needs to be validated when used in a setting different from where it was developed and also when it is used after translation. The aim of this study was to validate the Malayalam version of the ECOHIS (M-ECOHIS).
| Materials and Methods|| |
A cross-sectional design was used in the conduct of the study. The study was conducted across various government-run and private kindergartens in the capital city of Kerala, Thiruvananthapuram. Kerala is an Indian state located on the Malabar Coast of southwest India. Malayalam is Kerala's official language.
Healthy preschool children (aged 3-5 years) were included in the study. Informed consent was obtained from all the parents. The city was divided into four geographic regions. From each region, two wards were randomly selected. From the selected wards kindergartens were randomly selected. All children in the selected kindergartens were included in the study. The sampling procedure ensured representativeness from all parts of the city as well as from both high and low socioeconomic classes. The study consisted of a pilot study and a main study. The main study included the administration of the M-ECOHIS and a global oral health (GOH) question by the principal investigator. Also recorded were the demographic details of the children and their parents. Approval was obtained from the Institutional Ethical Committee of the Government Dental College, Thiruvananthapuram.
The M-ECOHIS was derived forward-backward translation process. The process consisted of several stages. First, the measure was translated from English into Malayalam by a bilingual (Malayalam and English speaking) person well versed in both the languages. Secondly, the Malayalam version was revised through a consultation process involving the principal investigator. The Malayalam version produced by this process was then back translated by another bilingual (Malayalam and English speaking) person well versed in both languages. This back translated English version was compared with the original English ECOHIS. Some changes were necessary for the questionnaire to be applicable to the culture and lifestyle in Thiruvananthapuram city. For example, a high number of mothers, who will be the main respondents, were housewives with no official job. Therefore, their children's illness would not result in taking time off from work (item 12), but might disrupt their normal daily activities at home or limit the time they usually spend with their other children. The question was therefore modified with the semantic, idiomatic, experimental, and conceptual equivalencies. A series of pre tests were conducted among colleagues and the respondent population to check the wording of items and comprehensibility of the tool. The translated version was then tested for face and content validity among experts.
A pilot study was conducted on a representative sample of 100 to determine the feasibility of the study and check for any further refinement of the tool if needed. The sample size for the main study was estimated at 270 after the pilot study. In order to compensate for the missing response, the sample size was increased by 10% and finalized at 300.
The data collection for the main study consisted of clinical assessment of dental caries and the administration of the following:
- A single item GOH question in Malayalam to assess the convergent validity of the ECOHIS,
- The 13 item M-ECOHIS questions, and
- An additional set of questions to assess the demographic characteristics like the child's age, sex, socioeconomic status, relationship of the respondent to the child, the mother's education, etc.
The data were collected by interviewing the parent or the primary care giver of the child. The M-ECOHIS was administered twice on a subsample of the parents (30) to assess test-retest reliability. Two raters administered the M-ECOHIS on another subsample (30) to assess the interobserver reliability. Following this, a type 3 clinical examination as per WHO specification was done with the help of a mouth mirror and explorer using good natural daylight. Dental caries was diagnosed using visual and tactile method, and the caries experience of the child was measured with the help of the defs index.
The data was tabulated using Microsoft Excel 2010 and analyzed using Statistical Package for Social Sciences (SPSS) version 11 software. Test-retest reliability and interobserver reliability of the tool were determined with the help of intraclass correlation coefficient (ICC). Internal consistency reliability was tested with the help of Cronbach's alpha test and item total correlation. Convergent validity was derived by correlating the scores of the GOH question and the 13-item ECOHIS questionnaire using the Spearman's correlation coefficient. The mean ECOHIS scores of the children with ECC were compared with the mean ECOHIS scores of children without ECC using the unpaired t-test to evaluate the discriminant validity of the ECOHIS. Both convergent validity and discriminant validity together helped determine the construct validity.
| Results|| |
Out of 300 invited parents, 297 participated in the study. Those who did not participate were unaware of the child's or family's events. The age of the children ranged from 3 to 5 years with the mean age being 4.11. 50.5%of the children were girls and 49.5% were boys. The respondents were the mothers of children. The M-ECOHIS scores ranged from 0 to 25.
Reliability and validity of the M-ECOHIS
The reliability of the M-ECOHIS was established using internal consistency reliability, test-retest reliability, and interobserver reliability. Construct validity was determined by assessing the convergent and discriminant validity.
Internal consistency reliability of the ECOHIS
The internal consistency reliability of the ECOHIS was evaluated by the Cronbach's alpha test and item total correlation. The Cronbach's alpha value for the scale was found to be 0.8793. The Cronbach's alpha ranged from 0.8602 to 0.882.
The item total correlations ranged from 0.2832 to 0.7617 and were all positive in nature and statistically significant. The lowest coefficients were related to "avoided smiling or laughing" (0.2832) followed by "avoided talking" (0.3772) [Table 1].
|Table 1: Internal consistency reliability of the M-ECOHIS: Item-total statistics|
Click here to view
Test-retest reliability and the interobserver reliability
The test-retest reliability of the ECOHIS was evaluated with the help of the intraclass correlation coefficient. Among the 30 subjects on whom the ECOHIS was administered for the second time 2 weeks after it was first administered, the intraclass correlation coefficient of the ECOHIS was 0.9457 (95% confidence interval, 0.9132-0.9702).
The interobserver reliability of the scale was also assessed with the help of the intraclass correlation coefficient. Two weeks after the scales was first administered, a second observer administered the ECOHIS on a subsample of 30 parents to assess the interobserver reliability. Among the 30 subjects, the intraclass correlation coefficient was found to be 0.9460 (95%confidence interval, 0.9143-0.9700).
The convergent validity of the scale was evaluated by correlating the scores of the single item GOH question and the 13-item ECOHIS questionnaire. Spearman's correlation coefficient was used to determine the correlation between the two scores. The Spearman's correlation coefficient of the ECOHIS scores and the GOH scores was found to be r = 0.725 (P = 0.001), which is a moderate and statistically significant correlation supporting the convergent validity of the scale [Table 2].
|Table 2: Convergent validity of the M-ECOHIS: The Spearman's correlation coefficient of the ECOHIS scores and the GOH scores|
Click here to view
Discriminant validity of the ECOHIS was evaluated by comparing the mean ECOHIS scores of the children without caries (defs = 0, the controls) with the mean ECOHIS scores of children with caries (defs ≥1, the cases). The mean ECOHIS score of the controls was 0.01, while the mean ECOHIS score of the cases was 5.42. The means were compared with the unpaired t-test (t = −9.781) and the difference was found to be statistically significant (P = 0.001) proving the discriminant validity of the ECOHIS [Table 3].
|Table 3: Discriminant validity of the M-ECOHIS: Comparison of the mean ECOHIS scores of cases and controls|
Click here to view
| Discussion|| |
The aim of this study was to develop and validate the M-ECOHIS by examining its internal consistency, test-retest reliability, interobserver reliability, convergent validity, and discriminant validity.
The M-ECOHIS was derived by the forward-backward translation process. The translated version was similar to the original version except for some changes that were necessary for the scale to be applicable to the culture in Kerala. A high number of mothers, who will be the main respondents, were housewives with no official jobs. Therefore, their children's illness would not result in their taking time off from work (item 12) but might disrupt their normal daily activities. The question was modified so as to include an option to report a disruption in the daily household activities due to the child's dental/oral condition.
The reliability of a test instrument concerns the extent to which the instrument yields the same result on repeated trials. Though unreliability is present to a certain extent, there will generally be a good deal of consistency in the results of a quality instrument gathered at different times. The tendency towards consistency found in repeated measurements is referred to as reliability. Internal consistency reliability, test-retest reliability, and interobserver reliability are all measures of reliability of a tool.
Internal consistency of the tool is measured by estimating the Cronbach's alpha coefficient and the item total correlation. Cronbach's alpha is a coefficient (a number between 0 and 1) that is used to rate the internal consistency or homogeneity or correlation of the items in a test. A good test is one that assesses different aspects of the trait being studied. If a test has a strong internal consistency it should show only moderate correlation among items (0.7-0.9).  If the correlations between items are too low, it is likely that they are measuring different traits, and therefore, should not all be included in a test that is supposed to measure one trait. If item correlations are too high, it is likely that some items are redundant and should be removed from the test. The results of this validation process indicated that Cronbach's alpha was 0.8793, which was above the recommended value of 0.70.  The Cronbach's alpha for the items ranged from 0.8602 to 0.882, indicating that all items possess good internal consistency.
The item-total correlation is one of the oldest and the most widely used methods to check the homogeneity of the items in a tool. As the name implies, it is the correlation of the individual item with the scale total omitting that item. The usual rule of the thumb is that an item should correlate with the total score above 0.2. Items with scores below 0.2 must be discarded.  All inter-item correlations were positive and above the recommended level of 0.2. The corrected item-total correlations were also well above the recommended level of 0.2.
Assessment instruments should be reproducible over time, that is, they should produce similar results on two or more administrations to the same individual, provided that the general clinical state has not been altered. The analysis of test-retest reliability suggests the adequate stability of the instrument. It is recommended that the interval between measurements be long enough to reduce the effects of memory and short enough to diminish the likelihood of systemic alterations. Although the definition of this interval is arbitrary, a period of 2-14 days is considered adequate. ,, The intraclass correlation coefficient (ICC) for test-retest reliability in the present study is 0.9457, which is close to that of the French ECOHIS (0.95)  and was higher than that of the original English version (0.84),  the Farsi ECOHIS (0.82),  the Chinese ECOHIS (0.64),  and the Turkish version (0.86).  The Brazilian ECOHIS  in comparison had a very high test-retest reliability of 0.99. The ICC gave evidence of satisfactory reproducibility of the M-ECOHIS. In the present study, the scale was administered to each parent by the principal investigator. Hence, the interobserver reliability was tested after the ECOHIS and was re administered on a subsample of 30 parents by another observer. The ICC for the interobserver reliability was 0.9460, indicating good reproducibility when administered by other investigators. The interobserver reliability is yet another evidence for the reliability of the M-ECOHIS.
Convergent validity is determined by hypothesizing and examining the overlap between two or more tests presumably measuring the same construct. In other words, convergent validity is used to evaluate the degree to which two or more measures that theoretically should be related to each other are, in fact, observed to be related to each other. In the present study, the correlation between the GOH question and the ECOHIS was estimated as a measure of convergent validity. The Spearman's correlation coefficient for ECOHIS and the GOH question was r = 0.725 (P = 0.001), which could be interpreted as moderate (Spearman's correlation coefficient of 0.6 and above),  supporting the convergent validity of the measure.
Discriminant validity of a tool examines the extent to which a measure correlates with measures of attributes that are different from the attribute the measure is intended to assess. Discriminant validity analysis is considered a useful method in the differentiation of groups that are known to be distinct. It is the ability of the tool to discriminate between those with the disease and those without the disease. In the present study, the discriminant validity was measured by comparing the mean ECOHIS scores (0.01) of children without ECC with the mean ECOHIS scores of children with ECC (5.42). This was done with the unpaired t-test. The difference in the means was found to be statistically significant (t = -9.781, P = 0.001). The M-ECOHIS was able to discriminate fairly well between children with ECC and children without ECC.
The M-ECOHIS is entirely based on the perceptions of parents/caregivers and their understanding of health and illness of the child. Parents of a child may have different views about their child's health from each other and from the child. The issue becomes more important when a child has been taken care of by different people, for example, mother and grandmother, in different periods of life.
There was some confusion among the parents about the options for the response to the ECOHIS. There was an overlap in the meaning of "hardly ever" and "occasionally"; as the verbal inference of both the options in Malayalam was perceived as the same by the parents. Using a simple Likert frequency type scale with three options - never, occasionally, and very often - would help to elicit a more meaningful conclusion.
The multistage sampling procedure ensured that a sample representing all parts of the city was selected in the main study. Selecting both the government run and the private kindergartens ensured that both the high and the low socioeconomic classes are included in the study.
The interview method of questionnaire administration used in this study helped to eliminate the "don't know" responses. The M-ECOHIS was developed and validated and demonstrated acceptable standards.
It will be valuable to test the M-ECOHIS in a rural population with differing levels of disease and also to evaluate the tools ability to identify changes in response to intervention.
The psychometric properties of the M-ECOHIS have been tested in this study using the classical test theory. The Rasch validation of the Chinese ECOHIS by Wong et al.,  has shown that the instrument has a range of difficulty levels across the items and the performance of item consistency. This study stresses the need to test the M-ECOHIS using the item response theory (Rasch model). Future research on the M-ECOHIS using the Rasch model will provide additional information that can be added to that obtained from the classical test theory. The Rasch model will also provide an estimate of a person's ability and item difficulty.
| Acknowledgment|| |
The authors would like to thank Pahel BT, Rozier RG, Slade GD for providing the ECOHIS tool.
| References|| |
Ayhan H, Suskan E, Yildirim S. The effect of nursing or rampant caries on height, body weight and head circumference. J Clin Pediat Dent 1996;20:209-12.
Thomas C, Primosch R. Changes in incremental weight and well-being of children with rampant caries following complete dental rehabilitation. Pediatr Dent 2002;24:109-13.
Low W, Tan S, Schwartz S. The effect of severe caries on the quality of life in young children. Pediatr Dent 1999;21:325-6.
Sartorius N. The Meanings of Health and its Promotion. Croat Med J 2006;47:662-4.
Pahel BT, Rozier RG, Slade GD. Parental perceptions of children′s oral health: The Early Childhood Oral Health Impact Scale (ECOHIS). Health Qual Life Outcomes 2007;5:6.
Nunnally JC, Bernstein IH. Psychometric Theory. 2 nd
ed. New York: McGraw-Hill, Inc; 1978; p. 253-68.
Kline P. A handbook of psychological testing. London: Routledge, 1 st
ed. 1993; p. 178.
Guyatt GH, Feeny DH, Patrick DL. Measuring health-related quality of life. Ann Intern Med 1993;118:622-9.
Shrout PE. Reliability. In: Tsuang Mt, Tohan M, Zhaner G.E.P (Eds), editor. Textbook in Psychiatry Epidemiology. New York: Wiley-Liss; 2 nd
ed. 1995. p. 213-27.
Streiner DL, Norman GR. Health measurement scales: A practical guide to their development and use. Oxford: Oxford University Press; 3 rd
ed. 2003, p. 187-202.
Li S, Veronneau J, Allison PJ. Validation of a French language version of the Early Childhood Oral Health Impact Scale (ECOHIS). Health Qual Life Outcomes 2008;6:9.
Jabarifar SE, Golkari A, Ijadi MH, Jafarzadeh M, Khadem P. Validation of a Farsi version of the early childhood oral health impact scale (F-ECOHIS). BMC Oral Health 2010;10:4.
Lee GH, McGrath C, Yiu CK, King NM. Translation and validation of a Chinese language version of the Early Childhood Oral Health Impact Scale (ECOHIS). Int J Paediatr Dent 2009;19:399-405.
Scarpelli AC, Oliveira BH, Tesch FC, Leão AT, Pordeus IA, Paiva SM. Psychometric properties of the Brazilian version of the Early Childhood Oral Health Impact Scale (B-ECOHIS). BMC Oral Health 2011;11:9.
Peker K, Uysal Ö, Bermek G. Cross-cultural adaptation and preliminary validation of the Turkish version of the Early Childhood Oral Health Impact Scale among 5-6- year-old children. Health Qual Life Outcomes 2011;9:118.
Cohen LK, Jago JD. Toward the formulation of a sociodental indicator. Int J Health Serv 1976;6:681-98.
Wong HM, McGrath CP, King NM. Rasch validation of the early childhood oral health impact scale. Community Dent Oral Epidemiol 2011;39:449-57.
[Table 1], [Table 2], [Table 3]