Secondary Logo

Journal Logo

Healthcare Economics, Policy and Organization

A Survey of Charge Sensitivity and Charge Awareness Among Intensive Care Unit Providers in a Large Academic Medical Center

Kingeter, Adam J. MD*; Shotwell, Matthew S. PhD; Parmley, C. Lee MD, JD, MMHC*; Pandharipande, Pratik P. MD, MSCI*; Buntin, Melinda B. PhD

Author Information
doi: 10.1213/ANE.0000000000003679


According to the most recent data reported by the National Center for Health Statistics, total national health care expenditures in 2015 totaled $3.2 trillion, accounting for 17.8% of the Gross Domestic Product, with hospital care constituting 32% of all health care expenditures.1 Significantly contributing to the high levels of health care spending are wasteful or unnecessary practices by providers, which account for an estimated one-third of all health care costs.2

As a means of curtailing wasteful practices and costs, the Centers for Medicare and Medicaid services has begun to move away from traditional fee-for-service payments, replacing them with payments that tie provider reimbursement to measures of quality and efficiency. These so-called value-based payments, particularly bundled and capitated payments, shift a large portion of the responsibility for costs of care onto providers, and put them at increased financial risk for providing care. Under these payment strategies, providers with little cost sensitivity or awareness are at risk of exceeding allocated payments and incurring uncompensated expenses.

Unfortunately, physician sensitivity to cost in clinical decision making is conflicting, and physician cost awareness is low.3–5 Hospital care involving the intensive care unit (ICU) is of particular note because ICU care represents an estimated 13.4% of all hospital care expenditures.6 To date, no data exist regarding the attitudes and financial awareness of providers in a multidisciplinary critical care environment in the United States. To this end, we aimed to characterize the charge consciousness and charge awareness of different provider types across several ICUs in a large academic medical center.


A 10-question multiple-choice survey was developed as part of the ICU Physician Awareness of the ICU Charge Environment (ICU-PRICE) study (NCT02476591), which was approved by the institutional review board, and the requirement for written informed consent was waived. The survey was sent before the implementation of the ICU-PRICE study intervention. The survey contained 2 questions regarding provider type and years at institution, 3 questions assessing the respondents’ charge sensitivity, and 5 questions assessing the respondents’ charge awareness. Charge awareness was assessed by asking providers to select the correct approximate charge out of 5 possibilities for 5 commonly ordered diagnostic tests and therapeutic treatments. We elected to use charge data obtained from the hospital master charge list as a surrogate for cost. Although imperfect, charge data are the most readily available to providers via the hospital master charge list, and thus, we hypothesized, the data with which providers are most likely to be familiar. We defined charge sensitivity by the degree to which providers agreed that controlling health care expenses was part of their responsibility, and they both knew and considered charges when prescribing medications or treatments for their patients; providers who strongly agreed to all 3 were considered to be most sensitive. We defined charge awareness by the number of correct charge estimates by each surveyed provider.

The survey was designed and administered in REDCap (supported by grant UL1 TR000445 from National Center for Advancing Translational Sciences/National Institutes of Health [NCATS/NIH]). Providers were sent an email explaining the study, the confidential and anonymous nature of the survey, and directing them to the online survey instrument. A repeat email was sent halfway through the 4-week study period, reminding respondents of the study and requesting their participation.

The survey was sent to all interns and residents in general surgery, internal medicine (IM), and anesthesiology, as well as all nurse practitioners (NP) and physician assistants (PA) who staff the cardiovascular, surgical, burn, neurosurgical, and medical ICUs. The attending IM, trauma/surgical, and anesthesiology intensivists who staff those units were invited as well, as were all IM, anesthesiology, and trauma/surgical critical care fellows. Respondents were asked to classify themselves by provider role as intern, resident, fellow, attending, or midlevel provider (NP or PA). For analysis, interns, residents, and fellows were grouped as physicians-in-training (PIT).

Proportional odds logistic regression (POLR) was used to evaluate associations between charge sensitivity (ordered categorical-dependent variable) and provider job title, adjusting for years of service (independent variables). We implemented a secondary analysis by additionally adjusting for “strong disagreement” with the “know charges” statement. The POLR method enables the analysis of ordered outcomes that might otherwise be dichotomized. For POLR, the odds ratio (OR) is interpreted identically for any binary grouping of the outcome that preserves its order. Thus, the OR is often described generically, for example, “the odds of greater agreement.” The validity of this assumption was checked graphically by comparing the estimated POLR OR with a series of ORs that were estimated separately for each binary grouping (Supplemental Digital Content 1, Figure 1, Binary logistic regression (LR) was used to quantify the associations between correct estimation of each type of charge and provider role, adjusting for years of service, and level of agreement with the “know charges” statement.

OR estimates are presented with 95% confidence intervals (CIs). Wald tests were used to evaluate the overall significance of independent variables. The Wilson score method was used to compute 95% CIs for proportions. CIs that failed to include the null value and P values <.05 were considered statistically significant.

No formal a priori statistical power or precision analysis was implemented. Nevertheless, this sample generated strong evidence about the associations between job title and cost awareness, as evidenced by the CIs presented below.


A total of 452 providers were invited to participate in the survey, with a response rate of 65.2% (294 individuals). PIT comprised 70% of respondents, while attending physicians and midlevel providers each comprised 15% of total responses. Among respondents, 20% reported having worked at the institution for <1 year, 55% reported having worked at the institution between 1 and 5 years, and 25% reported having worked at the institution >6 years.

An overwhelming majority (92.5% [95% CI, 89.0–95.0]) of respondents strongly or somewhat agreed that controlling expenses is part of their responsibility as a health care provider. Adjusting for years of experience at the hospital, attending physicians (POLR OR, 3.83; 95% CI, 1.52–9.66) and NP/PAs (POLR OR, 5.07; 95% CI, 2.30–11.16) were significantly more likely to agree that controlling expenses was part of their responsibility when compared to PIT (Figure 1; POLR P < .001). Among all respondents, 53.6% (95% CI, 47.9–59.1) strongly or somewhat agreed that they routinely consider charges when prescribing medications or diagnostic tests. After adjusting for years of experience, attending physicians (POLR OR, 2.17; 95% CI, 0.89–5.26) and NP/PAs (POLR OR, 3.72; 95% CI, 1.76–7.85) were more likely to agree that they consider the charges for medications and tests that they prescribe relative to PIT (Figure 1). However, 87.4% (95% CI, 83.2–90.8) of all respondents disagreed (somewhat or strongly) that they knew charges for most of the diagnostic tests and medications they prescribe for their patients. Attending physicians (POLR OR, 3.13; 95% CI, 1.33–7.36) and NP/PAs (POLR OR, 2.00; 95% CI, 0.98–4.08) were more likely to report greater knowledge of charges when compared to PIT, after accounting for years of experience. Providers who strongly disagreed with the “know charges” statement were significantly less likely to agree that they consider charges when prescribing (POLR OR, 0.62; 95% CI, 0.40–0.96) and significantly less likely to agree that controlling expenses was part of their responsibility as providers (POLR OR, 0.54; 95% CI, 0.33–0.86). There was no evidence that number of years of experience at our institution was independently associated with response to any of the charge-sensitivity questions after adjusting for provider job title.

Figure 1.
Figure 1.:
Percentage of respondents who agreed strongly or somewhat in response to 3 survey questions regarding perceived role in controlling expenses, knowledge of charges, and consideration of charges in clinical practice. Respondents were self-classified as attending physicians, nurse practitioners (NP)/physician assistants (PA) or physicians-in-training (PIT), which included interns, residents, and fellows.
Figure 2.
Figure 2.:
Provider selections of approximate charge for 3 commonly ordered laboratory tests. Estimates are shown for the entire cohort and grouped by provider type (physicians-in-training [PIT], nurse practitioners [NP]/physician assistants [PA], attending physician). There were no significant differences between provider types in terms of charge estimation. The correct charge for a complete blood count (CBC) was $100, with overestimates being $150 or $300, and underestimates being $15 or $50. The correct charge for a basic metabolic panel (BMP) was $150, with overestimates being $300 and underestimates being $15, $50, or $100. The correct charge for a serum potassium was $50, with overestimates being $100, $150, or $300, and underestimates being $15.

Averaging across all 5 surveyed items, just 35% (95% CI, 32.9–37.8) of provider selections of the approximate hospital charge were correct (Figure 2). There was no evidence that correct selection varied significantly among provider types (LR: P > .25 for all tests/diagnostics). In addition, level of agreement with the “know charges” statement was not significantly associated with correct estimation of charges (LR: P > .15 for all tests/diagnostics). Among incorrect estimates, more providers underestimated the correct charge for the surveyed item than overestimated (Figure 2). A more detailed description of provider estimates for each surveyed item is presented in Supplemental Digital Content 2, Figure 2,


This study is the first to demonstrate that the majority of ICU providers in a large academic medical center in the United States agree that controlling expenses is part of their responsibility as health care providers. This is consistent with results of previous surveys examining provider attitudes toward cost control.4,5,8,9 However, consistent with previous studies evaluating charge awareness in other providers, overall charge awareness in this cohort of American ICU providers was low.4,5,10,11 This lack of charge awareness limits the ability of providers to make value-based decisions for their patients. Nonetheless, a significant portion of providers attempt to incorporate charges into their clinical decision making. These results suggest that ICU providers are sensitive to their role in controlling health care expenses and receptive to incorporating financial data into their clinical decision making. The results of the ICU-PRICE study will determine whether offering ICU providers these data will lead to better stewardship of resources and decreased health care spending.


Name: Adam J. Kingeter, MD.

Contribution: This author helped with study conceptualization, design, data analysis, and manuscript preparation.

Name: Matthew S. Shotwell, PhD.

Contribution: This author helped analyze the data and prepare the manuscript.

Name: C. Lee Parmley, MD, JD, MMHC.

Contribution: This author helped design the study and prepare the manuscript.

Name: Pratik P. Pandharipande, MD, MSCI.

Contribution: This author helped design the study and prepare the manuscript.

Name: Melinda B. Buntin, PhD.

Contribution: This author helped design the study, analyze the data, and prepare the manuscript.

This manuscript was handled by: Nancy Borkowski, DBA, CPA, FACHE, FHFMA.


1. National Center for Health Statistics. Health, United States, 2015: With Special Feature on Racial and Ethnic Health Disparities. 2016.Hyattsville, MD: National Center for Health Statistics (US).
2. Berwick DM, Hackbarth AD. Eliminating waste in US health care. JAMA. 2012;307:1513–1516.
3. Tilburt JC, Wynia MK, Sheeler RD, et al. Views of US physicians about controlling health care costs. JAMA. 2013;310:380–388.
4. Nethathe GD. Cost awareness on the part of health professionals. S Afr Med J. 2015;105:618.
5. Hernu R, Cour M, de la Salle S, Robert D, Argaud L; Costs in French ICU Study Group. Cost awareness of physicians in intensive care units: a multicentric national study. Intensive Care Med. 2015;41:1402–1410.
6. Halpern NA, Pastores SM. Critical care medicine in the United States 2000-2005: an analysis of bed numbers, occupancy rates, payer mix, and costs. Crit Care Med. 2010;38:65–71.
7. Harrell FE. Regression Modeling Strategies: With Applications to Linear Models, Logistic Regression, and Survival Analysis. 2001.New York, NY: Springer.
8. Ginsburg ME, Kravitz RL, Sandberg WA. A survey of physician attitudes and practices concerning cost-effectiveness in patient care. West J Med. 2000;173:390–394.
9. Altomare I, Irwin B, Zafar SY, et al. Physician experience and attitudes toward addressing the cost of cancer care. J Oncol Pract. 2016;12:e281–e288, 247.
10. Allan GM, Lexchin J, Wiebe N. Physician awareness of drug cost: a systematic review. PLoS Med. 2007;4:e283.
11. Ryskina KL, Smith CD, Weissman A, et al. US internal medicine residents’ knowledge and practice of high-value care: a national survey. Acad Med. 2015;90:1373–1379.

Supplemental Digital Content

Copyright © 2018 International Anesthesia Research Society