A substantial increase in the use of anesthesia services for outpatient gastrointestinal (GI) procedures may represent low-value care.1,2 Some investigators have argued that this increase reflects increases in patient risk,3 whereas others state that it may come from physicians “upcoding” risk because anesthesia services are often covered only for patients for whom sedation or anesthesia carries a high risk of complications.4,5 A study was conducted to assess whether coding practices on claims for anesthesia services have changed with time.
We used the Truven MarketScan data from 2005 to 2013 to identify the use of anesthesia services during this period. We limited our analytic sample to 18.9% (1 001 841 of 5 313 979) of GI anesthesia service claims with a coded American Society of Anesthesiologists (ASA) classification, and excluded patients younger than 18 or older than 65 years and those without continuous enrollment in an employer-sponsored, non–capitation-based health insurance plan for the 6 months preceding the respective GI procedure. The GI procedures and anesthesia services were identified using the codes in Current Procedural Terminology, Fourth Edition (CPT-4), published by the American Medical Association.1
The outcome measure indicates whether a patient was coded as having a high (an ASA level of III or higher) or a low (an ASA level of I or II) anesthesia risk. We predicted a patient’s probability of being coded as having a high risk as a function of calendar year, age, sex, service setting, geographic region, insurance type, and procedure type, and a set of respiratory, cardiovascular, and other chronic conditions. Robust standard errors were used to account for patients with multiple GI procedures. The main results are presented as predicted probabilities adjusted for comorbid conditions and with patient-level characteristics held at their means. As a subanalysis, we added physician fixed effects into the model using data from 2010 to 2013 to test changes with time in coding practices of the same physicians.
This study was approved by the Institutional Review Board of the RAND Corporation.
The proportion of cases with ASA coding increased from 2.9% (23 345 of 812 513 cases) in 2005 to 13.2% (224 852 of 1 697 928 cases) in 2013. The proportion of patients coded as having a high risk increased from 11.6% (2589 of 22 297 patients) in 2005 to 18.9% (39 058 of 207 117 patients) in 2013. The main analysis included 849 745 GI procedures and the subanalysis contained 91 907 GI procedures and 1440 unique physicians. Patient characteristics are described in the Table.
The Figure shows that the predicted probability of being coded as having a high risk of anesthesia more than doubled for all conditions from 2005 to 2013, indicating potential upcoding. The probability for patients with sleep apnea, for example, increased from 8.8% in 2005 to 21.5% in 2011 and remained at 20.8% in 2013. A similar pattern was also found among patients without any chronic conditions.
In the subanalysis, the odds of patients with similar characteristics being coded as being at high risk in 2011 were approximately twice those in 2010, more than 3 times those in 2012, and about 5 times those in 2013, with all year-to-year changes found to be statistically significant.
Coding practices for anesthesia services changed with time, and there is evidence of potential upcoding of patient anesthesia risk. Our results cannot be explained by the severity of patients’ conditions. Neither can they be attributed to changes in the physician population, in that the changes in coding for anesthesia risk become more marked when the same physicians were examined over time. It also seems unlikely that the prevalence of risks that we cannot detect on the basis of claims data (eg, allergies to sedatives and airway abnormalities) would have more than doubled during the study period. A likely explanation for this change is therefore that physicians used their clinical discretion to systematically change coding practices because coding a patient as being at high risk in a claim ensures payment of the claim. Nevertheless, our study has limitations. We relied on claims data and may have missed conditions affecting anesthesia risk. We also may have underestimated potential upcoding because physicians became more likely to report comorbidities over time6 and we may have overadjusted for the severity of patients’ degrees of illness. Furthermore, only a minority of cases had their ASA level coded, and the proportion with a coded ASA level changed over time.
Corresponding Author: Hangsheng Liu, PhD, RAND Corporation, 20 Park Plaza, Ste 920, Boston, MA 02116 (hliu@rand.org).
Published Online: May 2, 2016. doi:10.1001/jamainternmed.2016.1244.
Author Contributions: Dr Liu and Ms Nie had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Nie, Mattke, Liu.
Acquisition, analysis, or interpretation of data: Nie, Predmore, Liu.
Drafting of the manuscript: Nie, Predmore, Liu.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Nie, Predmore, Liu.
Obtained funding: Liu.
Administrative, technical, or material support: Predmore.
Study supervision: Mattke, Liu.
Conflict of Interest Disclosures: None reported.
Funding/Support: This study was supported by Ethicon Endo-Surgery, Inc.
Role of the Funder/Sponsor: The sponsor had no role in design and conduct of the study; collection, management, analysis, and interpretation of the data; and preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication, except that they provided comments on a draft manuscript.
Additional Information: Sedasys is the manufacturer of the SEDASYS Computer-Assisted Personalized System. The system is intended to allow trained physician-led teams to deliver minimal to moderate sedation with propofol to patients at low risk of complications during colonoscopy and other procedures.
1.Liu
H, Waxman
DA, Main
R, Mattke
S. Utilization of anesthesia services during outpatient endoscopies and colonoscopies and associated spending in 2003-2009.
JAMA. 2012;307(11):1178-1184.
PubMedGoogle ScholarCrossref 2.Lichtenstein
DR, Jagannath
S, Baron
TH,
et al; Standards of Practice Committee of the American Society for Gastrointestinal Endoscopy. Sedation and anesthesia in GI endoscopy.
Gastrointest Endosc. 2008;68(5):815-826.
PubMedGoogle ScholarCrossref 3.Riviello
K. Why we need anesthesia providers, not robots or RNs, in GI settings pushing medications.
Gastroenterology & Endoscopy News. 2014;65:6.
Google Scholar 4.Fleisher
LA. Assessing the value of “discretionary” clinical care: the case of anesthesia services for endoscopy.
JAMA. 2012;307(11):1200-1201.
PubMedGoogle ScholarCrossref 5.Rex
DK. Effect of the Centers for Medicare & Medicaid Services policy about deep sedation on use of propofol.
Ann Intern Med. 2011;154(9):622-626.
PubMedGoogle ScholarCrossref 6.Vaughan-Sarrazin
MS, Lu
X, Cram
P. The impact of paradoxical comorbidities on risk-adjusted mortality of Medicare beneficiaries with cardiovascular disease.
Medicare Medicaid Res Rev. 2011;1(3):E1-E17.
PubMedGoogle ScholarCrossref