What explains DRG upcoding in neonatology? The roles of financial incentives and infant health
Section snippets
Introduction and background
In the last two decades many industrialized countries have introduced prospective payment systems for the reimbursement of hospital inpatient care based on so-called diagnosis-related groups (DRGs). By paying a flat amount conditional on patient characteristics to health care providers, such payment systems generally aim at a more efficient allocation of resources in health care (Ellis and McGuire, 1993). Payment according to DRGs limits hospitals’ incentives to provide unnecessary treatment
DRGs in German neonatology
Before 2003, neonatal care in German hospitals was reimbursed on a per diem basis. Since the introduction of the DRG system, reimbursement is based on the following case characteristics: birth weight (or admission weight), surgical (OR-) procedures, long-term artificial respiration, (severe) complications, 5-day and 28-day mortality. Birth weights are classified along eight threshold values: 600 g, 750 g, 875 g, 1,000 g, 1,250 g, 1,500 g, 2,000 g, and 2,500 g. Reimbursement changes substantially at
Data
We use two data sources in this study. The first are official German birth statistics 1996–2010, covering both the pre-DRG- and the DRG-period in German hospitals. These data include about 10 million births, of which some 748,000 or 7.4% were of low birth weight (<2,500 g). To illustrate, Fig. 2 shows the number and percentage of births, by birth weight category and period (before/after introduction of DRGs). Fig. 2 indicates a general trend towards a larger proportion of live births with low
Changes in the birth weight distribution over time
We begin by describing trends in the distribution of recorded birth weights from 1996 to 2010, and how changes in the distribution might be related to the introduction of DRGs. Basically, we show that since the introduction of DRGs, recorded birth weights that have made their way into official statistics have been systematically bent below birth weight thresholds that are relevant for reimbursement. If thresholds are irrelevant for reimbursement, however, there is no such change in the
Reimbursement differentials and upcoding
We now examine the hypothesis that upcoding is particularly prevalent at financially salient thresholds, i.e. where the financial benefits of upcoding are particularly large in absolute terms. To that end, we have computed the expected reimbursement difference between adjacent DRGs in terms of birth weight at each relevant threshold and for each year. This is the payment difference that would be expected if one made a naive forecast of the relative weights in the current year based on the
Newborn health status and upcoding
We now come to what we believe is the main innovation of our research. Even conditional on DRG classification criteria, newborns are in different health states. In this section we examine whether newborn health as measured by gestational age, APGAR scores, and early (≤7 days) neonatal death are systematically related to upcoding. The birth register data include only birth length as measure of infant health other than birth weight. Thus we now use complementary information collected for the
Summary and conclusion
In this paper, we show that since the introduction of DRGs in German neonatal care, birth weights around thresholds relevant for reimbursement are increasingly manipulated, i.e. systematically shifted below the thresholds. We estimate that, between 2003 and 2010, about 12,000 newborns with birth weights just above DRG thresholds have been recorded as having a birth weight below the threshold. At some particularly salient thresholds the number of infants with birth weights coded just below the
Acknowledgements
An earlier version of this paper has been circulated under the title: “First do no harm, then do not cheat: DRG upcoding in German neonatology”. We are grateful to the editor, two anonymous reviewers, participants at the conference on “Empirical Health Economics” at Ifo/mea/University of Munich, the PhD-Seminar on Health Economics and Policy, Grindelwald, the meeting of the Health Economics Section of the German Economic Association, especially our discussant Hans-Helmut König, and seminar
References (27)
- et al.
Apgar score and the risk of cause-specific infant mortality: a population-based cohort study
Lancet
(2014) - et al.
Estimates of the cost and length of stay changes that can be attributed to one-week increases in gestational age for premature infants
Early Human Development
(2006) - et al.
Supply induced demand for newborn treatment: evidence from Japan
Journal of Health Economics
(2014) - et al.
Medicare upcoding and hospital ownership
Journal of Health Economics
(2004) - et al.
Hospital ownership and cost and quality of care: is there a dime's worth of difference?
Journal of Health Economics
(2001) - et al.
Actuarial day-by-day survival rates of preterm infants admitted to neonatal intensive care in New South Wales and the Australian Capital Territory
Archives of Disease in Childhood. Fetal and Neonatal Edition
(2013) - et al.
Effect of the introduction of diagnosis related group systems on the distribution of admission weights in very low birth weight infants
Archives of Disease in Childhood. Fetal and Neonatal Edition
(2011) - et al.
Estimating marginal returns to medical care: evidence from at-risk newborns
Quarterly Journal of Economics
(2010) Basisfallwerte (Zahlbetrag) aller DRG-Krankenhäuser 2003–2012
(2012)- et al.
Saving babies? Revisiting the effect of very low birth weight classification
Quarterly Journal of Economics
(2011)
The contribution of preterm birth to infant mortality rates in the United States
Pediatrics
The continuing value of the APGAR score for the assessment of newborn infants
New England Journal of Medicine
The quality and completeness of birthweight and gestational age data in computerized birth files
American Journal of Public Health
Cited by (56)
Framing and subject pool effects in healthcare credence goods
2023, Journal of Behavioral and Experimental EconomicsPrice and income effects of hospital reimbursements
2022, Journal of Health EconomicsHospital response to a case-based payment scheme under regional global budget: The case of Guangzhou in China
2022, Social Science and MedicineCitation Excerpt :First, with the introduction of the DIP payment, hospitals may reclassify cases to earn higher points by providing more intensive treatments and/or up-coding. In the literature, up-coding is defined as the systematic classification of patients into higher severity levels, while increases in treatment intensity indicates a shift from conservative treatments to ones performing procedures (Jurges and Koberlein, 2015). Under the DIP scheme, the incentive for up-coding is straightforward in which providers could earn more points by simply manipulating the principal diagnosis code.
Analysis of root causes of problems affecting the quality of hospital administrative data: A systematic review and Ishikawa diagram
2021, International Journal of Medical InformaticsCitation Excerpt :Upcoding may result from different coding practices with varying legal implications, including coding comorbidities comprehensively to raise treatment costs; substituting the primary diagnosis by a secondary diagnosis; and adding comorbidities that are not documented. The case of upcoding through birth weight of newborns was documented in several studies [7,92,99]. This information is used to determine the DRG of a hospitalization, with hospitalizations of newborns documented with low birth weight yielding substantially higher payments.