ICMR Research & Innovation Scale needs nuanced credit grading

Sir,

The perspective article published in the August 2025 in the Indian Journal of Medical Research on the ICMR-Impact of Research and Innovation Scale (ICMR-IRIS)¹ asserts that newly coined Publication-Equivalent (PE) index by the Indian Council of Medical Research (ICMR) is an inclusive and simple research impact evaluation tool that obviates the need to depend on both the non-inclusive but easy to use publication and citation based methods practiced widely as well as the highly inclusive but cumbersome methods like the one recommended recently in a government report². This index has eight indicators, with the PE values ranging between 1 and 20 assigned to them. Here, it is pointed out that the relative values assigned appear unreasonable. For example, given that a paper published in any peer-reviewed indexed journal – irrespective of citation and impact factor (IF), which could be even 0 each – represents 1 PE, it seems unjustified that a paper receiving 100 citations or a paper in a 10 IF journal has been considered worth 2 PE only. Citation counts and IFs are not normally distributed. Showing a highly skewed distribution, citation counts, on a 0-100 scale, have been found to be less than 10 for a majority of papers, and 20 or more for a very small number of papers². Similarly, on a scale of 0-40, a vast majority of journals show an IF of <5, and only a few of ≥10³. Giving a PE of 2 for 0-20 citations or 0-3 IF, 5 for 20-100 citations or 3-10 IF, and 10 for ≥100 citations and ≥10 IF, for example, may encourage high-quality research.

Further, negative results, though valuable, could be difficult to publish. Once a project is approved, it is immaterial whether the results support the null or alternative hypothesis. Disincentivizing negative results by not considering unpublished but otherwise completed work in research evaluation may encourage unethical practices like data cherry-picking to force publications. Giving a PE of at least 1 for documented unpublished negative findings may help here.

Also, flatly giving a PE of 5 for national or international granted patents, irrespective of patent types, scope of invention, and number of countries, appears naïve. A patent can be granted for a process or a product. The product patents, representing the end result of an invention, are considered far more superior and prized forms of intellectual property (IP) than method patents. Similarly, an international patent can be granted in one or multiple countries. Also, it may take several years before a patent is granted. Taking a nuanced view of IP rights would help assign patents with appropriate PE scores. Otherwise, a process patent granted in a country for a single, minor claim will enjoy PE worth 5 indexed journal publications, and a product patent granted in multiple countries with several major claims will only be worth half of a limited-use health technology that has been given 10 on PE score.