Comparison of Cost Center–Specific vs Hospital-wide Cost-to-Charge Ratios for Operating Room Services at Various Hospital Types

Abstract

This analysis estimates differences associated with using general hospital-wide cost-to-charge ratios vs cost center–specific cost-to-charge ratios for operating room services and discusses whether this bias varies by hospital type.

With value-based payment programs, accurate measures of cost are necessary for physicians and hospitals to remain financially viable. Administrative databases, such as the National Inpatient Sample, rely on hospital cost-to-charge ratios (CCRs) to convert patient billing charges to estimated hospital costs.^1,2 However, using a single conversion factor on all patient charges within a hospital may be an oversimplification, particularly for surgical costs, and may lead to inaccurate estimates. We assessed whether the cost center–specific CCR for operating room services differs from the general hospital-wide CCR and whether this deviation varies by hospital type.

Methods

Fiscal year 2015 financial statements from acute care hospitals in California were analyzed from June to December 2018. Details of the statement structure, inclusion criteria, and exclusion criteria are described elsewhere.³ Cost-to-charge ratios were calculated within each hospital by dividing total costs by total charges for the hospital (hospital CCR) and within each cost center by dividing total costs by total charges for that cost center (cost center–specific CCR). We focused on the surgery and recovery cost center (surgery CCR), which included direct costs (eg, salaries and supplies) and indirect costs of managing the operating room and recovery room.³ The data for this study are publicly available and do not meet the definition of human participants research. As such, institutional review board approval was not indicated, which we confirmed with the institutional review board of the University of California, Los Angeles, before starting the project.

Within each hospital, differences between the cost center–specific CCR and the hospital CCR were calculated. We then pooled hospital differences and calculated the statewide median. Interpretation of this statewide median was the systematic bias of using the hospital CCR instead of the cost center–specific CCR. Finally, we compared surgery CCRs by hospital teaching status and ownership (ie, government, for-profit, and not-for-profit facility). Bivariate comparisons were made using nonparametric tests with statistical significance determined using 2-sided tests and α = .05.

Results

Of the 300 acute care hospitals reporting surgical expense information in 2015, 289 had plausible CCRs (between 0 and 1) and were included in this analysis. The median hospital CCR was 0.25 (interquartile range [IQR], 0.21-0.30) and the median surgery CCR was 0.19 (IQR, 0.13-0.26); the surgery CCR was lower than the hospital CCR (−24.4%, IQR, −40.3% to −7.4%; absolute difference, 5.8 percentage points [IQR, 1.8-9.9]) (Table 1). Cost centers with lower CCR than the hospital CCR included the clinical laboratory (median relative difference, −41.9% [IQR, −52.3% to −28.4%]), anesthesia (−71.3% [IQR, −80.7% to −45.2%]), and computed tomography (−74.0% [IQR, −79.1% to −68.2%]); cost centers with higher CCR than the hospital CCR included the medical/surgical intensive care unit (50.3% [IQR, 24.8%-77.2%]), medical/surgical unit (59.4% [IQR, 31.1%-96.4%]), and the blood bank (63.4% [IQR, 2.9%-162.0%]).

Table 1. Hospital-wide vs Cost Center–Specific CCRs for California Hospitals in Fiscal Year 2015^a.

Type of Facility	Sample Sizeb	Median (Interquartile Range)
		CCR	Deviation From Hospital CCR, %c
Hospital	289	0.25 (0.21 to 0.30)	NA
Surgery and recoveryd	289	0.19 (0.13 to 0.26)	−24.4 (−40.3 to −7.4)
Other cost centers
Computed tomography	266	0.07 (0.05 to 0.09)	−74.0 (−79.1 to −68.2)
Anesthesia	226	0.07 (0.04 to 0.14)	−71.3 (−80.7 to −45.2)
Clinical laboratory	289	0.15 (0.11 to 0.20)	−41.9 (−52.3 to −28.4)
Emergency department	264	0.18 (0.14 to 0.24)	−25.3 (−36.7 to −9.2)
Drugs sold to patients	289	0.21 (0.15 to 0.29)	−16.7 (−35.4 to 3.7)
Supplies sold to patients (eg, implants)	238	0.33 (0.23 to 0.47)	32.0 (−6.5 to 78.7)
Medical/surgical intensive care unit	248	0.35 (0.29 to 0.45)	50.3 (24.8 to 77.2)
Medical/surgical unit	270	0.39 (0.30 to 0.50)	59.4 (31.1 to 96.4)
Blood bank	196	0.45 (0.29 to 0.64)	63.4 (2.9 to 162.0)

Abbreviation: CCR, cost-to-charge ratio.

^aAll comparisons between the hospital-wide CCR and cost center–specific CCR had a P value <.001 using the Wilcoxon matched-pairs signed rank test.

^bSample sizes vary across cost centers based on the number of facilities reporting each cost center.

^cCalculated as ([cost center–specific CCR − hospital-wide CCR]/hospital-wide CCR).

^dIncludes inpatient operating rooms that can be used for inpatient or outpatient procedures.

The median surgery CCRs were similar at teaching and nonteaching hospitals (0.21 [IQR, 0.16 to 0.33] vs 0.19 [IQR, 0.13 to 0.25]; P = .12) and were higher at government-owned facilities (0.26 [IQR, 0.22 to 0.40]) compared with not-for-profit (0.19 [IQR, 0.14 to 0.24]) or for-profit facilities (0.16 [IQR, 0.11 to 0.22]) (P < .001) (Table 2). Surgery CCRs deviated less from hospital CCRs at government-owned (−18.2 [IQR, −39.3 to 13.6]) and for-profit facilities (−18.3 [IQR, −33.6 to 7.5]) compared with not-for-profit facilities (−27.8 [IQR, −41.7 to −13.0]) (P = .002).

Table 2. Surgery and Recovery CCRs Stratified by Hospital Status and Ownership Status for California Hospitals in Fiscal Year 2015.

Variable	Surgery and Recovery		Deviation From Hospitala
	Median CCR (IQR)	P Valueb	Median CCR (IQR), %	P Valueb
Hospital status
Teaching (n = 30)	0.21 (0.16 to 0.33)	.12	−24.0 (−41.0 to −5.6)	>.99
Nonteaching (n = 259)	0.19 (0.13 to 0.25)		−24.4 (−39.8 to −7.4)
Ownership status
Government (n = 44)c	0.26 (0.22 to 0.40)	<.001	−18.2 (−39.3 to 13.6)	.002
For-profit (n = 73)	0.16 (0.11 to 0.22)		−18.3 (−33.6 to 7.5)
Not-for-profit (n = 172)	0.19 (0.14 to 0.24)		−27.8 (−41.7 to −13.0)

Abbreviation: CCR, cost-to-charge ratio.

^aCalculated as ([cost center–specific CCR − hospital-wide CCR]/hospital-wide CCR).

^bP values were generated using Wilcoxon rank sum and Kruskal-Wallis equality of populations rank tests

^cIncludes state-, city-, and district-owned facilities.

Discussion

Mean hospital charges were 4 times higher than hospital costs, but this difference was approximately 25% higher for the surgery cost center. Therefore, the use of hospital CCRs may overestimate costs for surgical patients. For example, an $8000 charge for an outpatient cholecystectomy procedure would be estimated as a $2000 cost using a hospital CCR when the actual cost is closer to $1500 using the surgery CCR. This bias would be exaggerated for patients with short or negligible inpatient stays and may be especially problematic when comparing medical and surgical therapies. There may be an additional bias of using hospital CCRs when comparing hospitals by ownership, overestimating surgical costs for not-for-profit hospitals more than for government-owned facilities and for-profit facilities.

This study has several limitations. First, use of cost-center CCRs improved the accuracy of cost estimates compared with hospital CCRs; however, alternative approaches, such as time-driven activity-based costing are more accurate and actionable.⁴ Second, our data came from a single state, which limited generalizability; however, California is a large and diverse state with per capita health care spending near the nation’s mean.⁵

Conclusions

Clinicians and payers will increasingly rely on accurate measures of cost to make value-based treatment decisions and to ensure financial solvency. Cost-center specific CCRs can be generated⁶ and our findings suggest they should be used to provide more accurate measures of the cost of surgical care.