Pure and simple “truths” from administrative data

May 8th, 2012

Recent editorial in JAMA discusses an original research using administrative data[1] which appears in the same edition. The authors underscore the difficulty to find “pure and simple” truths due to inherent limitations from such source. They explain four factors: new guideline, new diagnosis code, new diagnostic test and lack of prognostic indicators which should be cautiously taken into consideration by researchers, before deriving conclusion.

While those factors are technically correct, however the authors forget to mention the long-standing issue related to the quality of claim data: fraud, waste and abuse. Indeed, it is part of systemic problem in US health care, causing careless spending between $82 billion and $272 billion in 2011[2]. Cases of health care fraud and abuse turn into more complexes, involving different types of health care professionals, sophisticated modus operandi and even crime organized. US government has launched Health Care Fraud Prevention and Enforcement Action Team (HEAT) to encounter. Beside via routine mechanism, fraud investigations are usually started following any anomalies detected by data analysis and algorithm[3]. From this point of view, performing low cost research (using administrative data) could serve to give feedback and provide early sign to save billion dollar healthcare spending.

Case-by-case analysis is needed to ascertain whether such research is more beneficial for clinical decision making or health system correction. Evidence showed that while coding accuracy improved over years, however variability between diagnosis groups remains problem[4]. Any research combining administrative data with other source (for example: with a disease specific clinical registry) will obviously ensure diagnosis validity. If it is not the case, any anomalies from administrative data still reflect the pure and simple “truths” of health systems.

What are the lessons for Indonesia? Since our government has committed to enact BPJS by 2014, we should anticipate the growing accumulated claim data transmitted, processed and stored within the database of that agency. After each claim processed and the money reimbursed, what should we do, then? It is on the hands of researchers to study and examine healthcare utilization, effectiveness of therapy, epidemiological description of such disease and many other things. In this point of view, I believe that if we start to think about the concept of administrative data comprehensively, from technical aspect until meaningful use, hopefully this country will gain much from those data to improve either clinical decision making or health systems. Even if the phenomenon of GIGOLO (garbage in, garbage out, low output) appears, we still have opportunity to use this as the early sign to perform more comprehensive way to improve the quality.

References

  1. Sarrazin MS, Rosenthal GE. Finding pure and simple truths with administrative data. JAMA 2012;307(13):1433-5.
  2. Berwick DM, Hackbarth AD. Eliminating waste in US health care. JAMA 2012;307(14):1513-6.
  3. Dube JF. Fraud in health care and organized crime. Med Health R I 2011;94(9):268-9.
  4. Fisher ES, Whaley FS, Krushat WM, et al. The accuracy of Medicare’s hospital claims data: progress has been made, but problems remain. Am J Public Health 1992;82(2):243-8.

Comments RSS

Leave a Reply

Name (required)

Email (required)

Website

Speak your mind

- online valtrex order - buy asacol - spy a cell phone - generic levitra -