Minnesota's Mental Health Crisis: Our Methodology

Newly released data help examine the financial side of mental health care and how it can limit access to services.

The Star Tribune recently analyzed published hospital billing data for compliance with federal regulations. The regulations are intended to increase price transparency for the consumer, but the first step requires hospitals to publish their data in a machine-readable format that people rarely use or see. We turned to Turquoise Health, a San Diego-based company that collects and normalizes information, to help us optimize our analysis.

For comparison, the Star Tribune selected the overall treatment category based on Minnesota's federal removal data in 2017, the most recent year available. The federal government uses these categories to classify hospital inpatient cases for Medicare payment decisions. Many private health insurers also use this payment method.

Among more than 700 categories of care, the Star Tribune selected the 20 most common, excluding those related to childbirth.

The analysis calculated the average payment for more than two dozen health plans at 88 hospitals in the state, a group that includes hospitals in Allina, CentraCare, Essentia Health, M Health Fairview, Mayo and Sanford Health Systems.

The average payment is adjusted according to the average length of stay. Commercial health plan, Medicaid and Medicare Advantage prices are compared to Medicare reference prices calculated by Turquoise Health.

The maternity category is excluded because Medicare does not pay for multiple births, so researchers question whether Medicare prices provide a meaningful measure.

More than 700 treatment categories are recognized as Medicare severity-related diagnostic groups. While there are other ways to pay with their diagnosis-based codes, these MS-DRG codes seem to be the most common type when hospitals bill hospital insurance, said Morgan Henderson, senior data scientist at the University's Hilltop Institute. of Maryland. Baltimore County.

Henderson conducted research on billing data published by hospitals across the country. Initial reports from hospitals in Minnesota and other states were criticized as incomplete and inconsistent, but the Centers for Medicare & Medicaid Services (CMS) says quality has improved significantly.

Each machine-readable file published by a hospital has thousands of line items. A file released last year by the University of Minnesota Medical Center, for example, contains about 34,000 lines of data, including codes for three main types of reimbursement: hospital service codes. professional health care service codes; and equipment codes used for maintenance.

The Star Tribune focuses on one part of the first group: hospital payment rates. The journal selected the 20 most common non-pregnant MS-DRGs in Minnesota based on federal data from the Agency for Healthcare Research and Quality (AHRQ).

The 20 categories accounted for about 27% of all Minnesota hospital discharges in 2017, the most recent year available when the Star Tribune began its investigation.

In March, the paper analyzed data from the Minnesota Hospital Association (MHA) and found that 15 of 20 MS-DRGs were also in the top 20 in 2021; the other five categories continue to rank in the top 10% of all MS-DRG codes.

The Star Tribune exported data from Turquoise Health in the fourth quarter of 2022 and again in the first quarter of 2023 and found very similar results in different time periods. Minor changes are expected as Turquoise Health continues to update its database.

The figures published here reflect the average calculated from the February 2023 data export.

There are different calculations of average length of existence (ALOS). The Star Tribune consulted with billing experts and researchers before choosing the arithmetic mean length of stay presented by CMS for 2022 to calculate the per diem amount.

Although there are arguments for using the geometric mean length of stay for CMS or LOS calculations in AHRQ (2017) or MHA (2021), most experts support the use of the CMS arithmetic mean. For mental patients, the arithmetic mean length of stay in CMS (8.8 days) is greater than the geometric mean length of stay in the body (5.9 days); it is also shorter than the AHRQ (9.7 days) and MHA (10 days) LOS estimates. So the figure chosen by the Star Tribune was neither the highest nor the lowest available.

Turquoise Health calculates the Medicare reference price for all MS-DRGs. To see how commercial health plan, Medicaid and Medicare Advantage pay for patients with psychosis compared with payments for other patient groups, the Star Tribune calculated the ratios for each of the 20 MS-DRGs.

On average, commercial health plans pay between 168% and 194% of the Medicare reference price, depending on the category. For psychosis, commercial plans pay 176% of Medicare rates.

Medicaid health plans pay between 103% and 122% of the Medicare reference price, depending on the category. For psychosis, they pay 113% of Medicare fees.

In all four payer categories (commercial standard, Medicaid, Medicare Advantage, and Medicare), average daily payments for psychosis were lower than all other treatment categories.

Turquoise Health's data includes out-of-pocket payments that reimburse the hospital for cases that require the hospital to use more resources. Figures published by the Star Tribune include outliers. Excluding the outlier, the mean changes only slightly, from 0% to -3.2% according to MS-DRG.

Here are 20 MS-DRGs from the Star Tribune analysis:

• 65. Intracranial hemorrhage or brain infection with CC or TPA within 24 hours
• 101. Seizures WITHOUT VFD
• 189. PULMONARY EDEMA AND RESPIRATORY FAILURE
• 190. CHRONIC OBSTRUCTIVE LUNG DISEASES with PKS.
• 193: PNEUMONIA AND PLEIRIS WITH PKS
• 194. SIMPLE PUMP AND WITH PLERIS CC
• 247. STENT WITH SELECTIVE DRUGS WITHOUT MCC.
• 291. HEART FAILURE and association with PKS
• 330: SMALL AND BIG PROCESS with CC
• 378. GASTROINTESTINAL HEMOLOGY CK
• 392. ESOPHAGITIS, GASTROENTERT AND OTHER GENERAL DISORDERS WITHOUT MCI
• 470. HEAD AND KNEE REPLACEMENT OR LOWER LIMB REPLACEMENT WITHOUT VFD
• 603. CELLULITE WITHOUT VFD
• 641. OTHER NUTRITIONAL, METABOLISM, FLUID AND ELECTROLYTE DISORDERS WITHOUT MCC
• 683. Renal failure with CC
• 690 Kidney and urinary tract infection WITHOUT PKS
• 871. SEVERE SEPTIMIA OR SEPSIS WITHOUT MV > 96 HOURS WITH PKS.
• 872. SEPTIMIA OR SEVERE SEPTIMIA WITHOUT MV > 96 HOURS WITHOUT MCC
• 885: PSYCHOSIS
• 897. ALCOHOL, DRUG ABUSE OR ADDICTION WITHOUT REHABILITATION WITHOUT ICU

The Mental Health Roundtable was supported by Call to Mind, MPR's mental health initiative