Response to questionable usefulness of the everyday abilities scale of India to screen for dementia.

Sir,

This is in response to correspondence entitled, “Questionable usefulness of the everyday abilities scale of India (EASI) to screen for dementia” regarding article entitled, “Feasibility of using everyday abilities scale of India as alternative to mental state examination as a screen in two-phase survey estimating the prevalence of dementia in largely illiterate Indian population” published in Indian Journal of Psychiatry.[1] We appreciated the effort taken by the reader in going through the contents of this article and responding with a correspondence. Before responding to the queries raised by the reader, we would like to revisit the stated aim of the study. The purpose of the post hoc analysis on the data obtained from a previous study was to explore the feasibility of using EASI as an alternative to Mini Mental State Examination in Hindi (HMSE) and its modifications as the first phase in two-phase survey to estimate the prevalence of dementia in largely, illiterate Indian population. Now coming to the use of Pearson's correlation instead of Spearman's in our study. The reader is reminded of the fact that the Pearson's correlation evaluates the linear relationship between two continuous variables while the Spearman's correlation evaluates the monotonic relationship between two continuous or ordinal variables and the Spearman's correlation coefficient is based on the ranked values for each variable rather than the raw data.[2] In the post hoc analysis performed by us, as stated above, we were dealing with raw data and ranked data. Importantly in a relationship wherein one variable decreases when the other increases, but the amount is not consistent, the Pearson's correlation coefficient is negative but > −1 while Spearman's correlation is −1, and as correctly pointed out by the reader, correlation values of −1 or 1 imply an exact linear relationship. The fact that while Pearson's correlation coefficients measure linear relationships and Spearman's correlation coefficients measure monotonic relationships, a meaningful relationship can exist even if the correlation coefficients are 0.[2] Examination of a scatter plot is therefore useful to determine the form of the relationship between two variables as has been shown by us in our article.[2]

This takes us to the second query, regarding not using receiver operating characteristic (ROC) curve. We would like to remind the reader that the ROC curve, no doubt, is an effective method of evaluating the performance of diagnostic tests;[3] here, in our study, we are purely dealing with two screening tests, and at no point in time, in our paper did, we talk of evaluating the usefulness of EASI as a diagnostic test. The basic idea behind the post hoc analysis was to evaluate the feasibility of using EASI as an alternative to HMSE and its modifications as it is easy to navigate through EASI in comparison to MMSE and therefore may be a useful tool in the first phase (screening phase) of two-phase survey to estimate the prevalence of dementia in largely, illiterate Indian population.

Regarding the concern of the reader on information about the sample on which the presented data are based, the details could not be included in view of the space of the article and were duly referenced.[4] The reader is requested to go through the reference for the detailed information.

We hope that we have been able to satisfy the reader through our response.

Finally, we express our regret for not being able to proof check the article thoroughly for the printing error on correlation coefficient as expressed by the reader in the final note. The value should be read as −0.252 and not as −2.52.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.