Descriptions and Changes in Charge Master Pricings of US Children’s Hospitals From 2019 to 2021

A list of hospitals was obtained by personal correspondence from a pediatric hospitalist who conducted an unpublished study that aimed to identify all hospitals across the United States that offered inpatient pediatric care. This list of hospitals was compiled between 2015 and 2017 and was assembled by calling every hospital listed (∼7600) in the American Hospital Directory database. When a hospital was called, the hospitalist asked the operator to be directed to the pediatric unit. If the hospital did not admit to pediatrics, no further calls were made. If the hospital had a pediatric unit, 3 attempts were made to get in contact with someone on the unit to determine their staffing model for the unit.

Hospital charge masters were searched for online using keywords including the hospital name, “charge master,” “charge description master,” “costs,” or “prices.” The 2019 hospital charge masters were searched for between July 2019 and December 2019. The 2021 hospital charge masters were searched for between October 2021 and December 2021.

CBC, CMP, and chest radiograph charge master prices were searched by using the find function in Microsoft Excel. CBC was searched for using keywords “CBC”, “blood count,” or the Current Procedural Terminology (CPT) code 85025. CMP was searched for using keywords “CMP,” “metabolic panel,” “chemistry panel,” or the CPT code 80053. Chest radiograph was searched for using keywords “CXR,” “chest x-ray,” or the CPT code 71045. Outside of these search terms, data collectors were also given the freedom to truncate search terms to more comprehensively search the charge master (ie, search for “complete,” “metabolic,” or “chest x”).

Descriptive statistics were used to characterize charge master prices of the 3 tests including the range, median (interquartile range [IQR]), and mean (SD). Coefficient of variation (CV) was calculated (SD divided by the mean) to standardize the variability comparisons between tests. The CV shows the extent of variability compared, with the mean with larger CV indicating larger dispersion. To investigate variability of prices within hospitals in close geographic proximity, we examined city-level variation for large metropolitan areas with multiple hospitals within. The top 6 cities with the most hospitals in our cohort were analyzed: Baltimore, California Bay Area, Chicago, Dallas-Fort Worth, Los Angeles, and New York City. To compare the differences in charges between 2019 and 2021, paired t test was used for hospital charge masters that were available in both years. For this comparison, 2021 prices were deflated to 2018 charges based off the US consumer price index: medical care index.

For hospitals that had complete datasets for all 3 tests and both years, z scores were calculated and organized into a heatmap to visually display correlations between charge master pricing by hospital. Correlation coefficients were also calculated between tests. P values less than .05 were considered statistically significant. Stata v15 was used for all statistics.

A total of 847 hospitals were identified, with all 50 states represented. There were 728 (86%) hospital charge masters found in 2019 and 747 (88%) found in 2021. There were 664 (78%) hospitals with charge masters obtained for both years.

CBC prices ranged from $5 from $1037 with median (IQR) of $110 (63, 177) in 2019, and $5 to $1085 with median (IQR) of $114 (65, 176) in 2021. CV was 92 in 2019 compared with 96 in 2021. Of the 633 hospitals with CBC charge prices available for both years, 94 (15%) had the same CBC price, 212 (33%) had decreased prices in 2021 compared with 2019, and 327 (52%) had increased prices in 2021 compared with 2019 (Table 1). After adjustment for inflation, there was no difference in average prices between 2019 and 2021 ($148 vs. $143, P = .17).

CMP prices ranged from $10 to $2440 with median (IQR) of $230 (102, 414) in 2019, and $6 to $2746 with median (IQR) of $246 (106, 444) in 2021. CV was 108 in 2019 compared with 107 in 2021. Of the 621 hospitals with CMP charge master prices available both years, 105 (17%) had the same CMP price, 168 (27%) had decreased prices in 2021 compared with 2019, and 347 (56%) had increased prices in 2021 compared with 2019 (Table 1). There was no difference in average prices between 2019 and 2021 ($347 vs. $331, P = .75).

Chest radiograph prices ranged from $20 to $1900 with median (IQR) of $301 (216, 456) in 2019 and $30 to $5800 with median (IQR) of $353 (242, 536) in 2021. CV increased from 65 in 2019 to 91 in 2021. Of the 647 hospitals with chest radiograph charge master prices available both years, 98 (15%) had the same chest radiograph price, 123 (19%) had decreased prices in 2021 compared with 2019, and 426 (66%) had increased prices in 2021 compared with 2019 (Table 1). Average prices were higher in 2021 compared with 2019 ($392 vs. $365, P < .0006).

Descriptive results for the city level variation of the 6 large metropolitan areas are shown in Table 2 and graphically depicted in Figs 1–3. We note that Baltimore consistently had the lowest median prices compared with the other 5 cities. When comparing variability changes from 2019 to 2021, we note that although the Bay Area and New York City had decreased CVs for all 3 tests, the other cities had inconsistent changes.

There were 595 hospitals with complete data for all 3 tests in both years. Correlation between tests and years varied between weakly positive to strongly positive (0.36-0.99) (Table 3). A heatmap of hospitals and tests with their z scores suggest correlations within hospitals in charge master testing prices (ie, a hospital who has a high z score value for 1 test is likely to have high z score values for other tests and vice versa (Fig 4).

We find that tremendous variability in charge master pricing continues to exist among US children’s hospitals 2 years after the mandated publication of hospital charge masters in 2019. This variability persisted even in metropolitan areas with many hospitals within proximity. This suggests that mandated posting of charge master prices has had little effect on exerting competitive pricing pressures theorized to happen with greater price transparency.

The variability of charge master prices has been demonstrated in previous studies that have looked at various procedures and tests such as knee or hip replacements, cholesterol tests, lower back magnetic resonance imaging scans, and total knee arthroplasty.^7,8  Our study demonstrates that there may be patterns to this variation. We find positive correlations between tests within hospitals, which suggests that hospitals with high charge master prices for 1 test likely have high charge master prices for other tests. This implies that there are a group of hospitals that consistently have higher charge master prices than their peers, as well as a group of hospitals that consistently have lower charge master prices than their peers. Future research into the differing characteristics of these hospitals would be informative. Charge master prices likely differ based on numerous factors including teaching status, size of the hospital, severity of illness of patients, quality metrics, profit status, overall payer mix, and location.⁹  On location, we note that Baltimore consistently had the lowest variability in city level analysis. Maryland is the only state in the United States where hospital payment regulation is set at a state level. The effects of this regulation continue to be studied.^10–12 

We question whether the current format of posting charge masters is truly increasing pricing transparency for patients. We found 3 themes of difficulty when collecting charge master prices from the hospital websites. First, we found difficulty finding the charge masters in the first place. Of the 847 hospitals, approximately 12% to 14% did not have their charge masters available. This is a much lower number than found in another study that looked at all US hospitals registered with CMS in 2021 that found that 86% of hospitals was not adherent to having a machine-readable file displayed.¹³  Our study results were similar to another study looking at a cross-section of hospitals in Pennsylvania that showed about 20% noncompliance,¹⁴  and 6% noncompliance from US News & World Report Honor Roll hospitals.⁸  Charge masters were often not clearly listed on the Web site, and it often took several clicks with the link embedded deep within the site. In their study, Mullens et al found that the mean number of clicks to access charge master was 3.7 (SD, 1.3; range, 1–8 clicks).¹⁴  There is no standardization among hospital Web sites where a patient would be able to easily access the charge master. Second, for some hospitals, we found difficulty downloading the charge master. The CMS mandate requires that the charge masters are posted in “machine-readable” files. Although most files were uploaded in .csv format that programs such as Microsoft Excel could open, some files were uploaded in .json format that either required uncommon third-party software to open or was downloadable as a text file that was not interpretable. Other charge masters simply gave error messages when clicking on the link to download or the files were too large to download. Finally, we had difficulty simply reading and finding the appropriate test in the charge master. There is an overwhelming number of fields in a charge master that is filled with medical jargon. Previous studies have shown that on average there are 41 000 number of fields in a charge master.⁸  There is no standardization in terminology or consistency with displaying items such as CPT codes among hospitals. It is unrealistic for an average patient with no medical knowledge to navigate different hospital charge masters to find the cost of for example, a chest radiograph when that test could be uploaded as any permutation of the words and abbreviations of chest radiograph (e.g., chest 1view, chest x-ray).

Our results should be interpreted with considerations of our limitations. First, the charge master prices collected for this study are only reflective of the prices at the time of collection. Hospitals may have changed the prices within the collection period and may also not be reflective of current prices. Second, we have no pre-2019 data on hospitals’ charge master prices before the mandate and so we cannot make any conclusions on the effect of the mandate on initial charge master price changes that may have occurred before our data collection. We also did not collect or compare charges based off the hospital’s display of “shoppable services” or payer-specific negotiated charges.¹⁵  Payer-specific negotiated rates may be more relevant for insured patients. These were inconsistently uploaded by hospitals and the number of insurance policies made comparisons difficult. Price estimator tools may be more consumer friendly than actual charge masters and may be more realistic means of comparing prices than searching the charge master. The most recent CMS rule mandates posting of these “shoppable services,” and so future research may be able to determine the effect of this intervention.

Mandated online publishing of hospital charge masters has not led to decreased charge master prices or decreased variability in three commonly ordered tests at pediatric hospitals over a 2-year span after the CMS mandate. Future studies should examine the effects of the availability of patient-centered price transparency tools on overall trends in health care costs and spending.

We thank Dr Clota Snow for providing the list of pediatric hospitals across the United States. We also thank Drs Youxin Xiong, Emily Ruseski, Tiffany Robles, Priscila Olague, Laura Moore, Mark Knackstedt, Tuong Phan, and Farhana Khaja for their help in data collection.