Hospital payments for COVID-19 inpatient treatment: An analysis of inpatient claims data
Our analysis of inpatient claims data shows large variations in hospital payments for COVID-19 inpatient treatment in 2020.
Medicare Advantage organizations (MAOs) offer Medicare beneficiaries a bundle of services that include Medicare Part A (hospital), Part B (medical), and usually Part D (prescription drug) coverages. Many MAOs offer supplemental benefits that are not covered under traditional fee-for-service Medicare. Dental coverage is a common supplemental benefit that may be offered as a mandatory supplemental benefit (MSB), where coverage is embedded in the Medicare Advantage (MA) plan, or an optional supplemental benefit (OSB), where consumers choose to add supplemental coverage.
MAOs choosing to participate in the Medicare Advantage Value-Based Insurance Design (VBID) Model may target benefits to enrollees based on chronic conditions and/or socioeconomic characteristics. MA VBID is a model promulgated by the Centers for Medicare and Medicaid Services (CMS) to test MA health plan innovations designed to reduce Medicare program expenditures, enhance the quality of care, and improve the coordination and efficiency of healthcare delivery.1 The targeted supplemental benefits may be primarily health-related, like dental, or non-primarily health-related (transportation, food assistance, etc.). In 2021, 19 MAOs covering 4.6 million enrollees (approximately 20% of MA enrollees) are participating in the MA VBID Model test.
In this paper, we used calendar year 2018 MA administrative claim data for enrollees aged 65 and older enrolled in plans with embedded dental coverage to determine the types of dental services used and the dental per member per month (PMPM) allowed costs, as well as to study correlations between medical and dental utilization. We segmented the studied population by chronic health conditions, by risk score, and by those with and without social determinant of health diagnoses. We also analyzed individuals we labeled high utilizers—members who seek significantly more health services than others with similar risk scores—to see if they use more dental services than average as well.
This information can be used as a baseline for changes that may come from the MA VBID Model in 2021.
Diabetes, cardiovascular disease (CVD), and chronic obstructive pulmonary disease (COPD) are chronic conditions for which many clinical studies have suggested connections between disease state and oral health. Examples of such studies include:
The table in Figure 1 shows that only a small portion of MA enrollees with embedded dental benefits incurred any dental services in 2018. We speculate that enrollees may be unaware of the benefits because they are embedded in the MA plan. Enrollees may also choose to pay out of pocket for dental benefits if MA dental networks do not include their dentists of choice. Also, the dental procedures covered by MA dental benefits vary and may not cover the specific services an individual is seeking.
|Percentage of Enrollees Utilizing Dental Services||All Enrollees||With CVD||With COPD||With Diabetes|
|No Dental Claims||88%||91%||92%||90%|
|Preventive Dental Claims||11%||8%||7%||9%|
|Other Dental Claims (not preventive)||1%||1%||1%||1%|
Note: Preventive claims include cleanings, oral exams, and periodontal cleanings.
We also summarized allowed dental and medical PMPM costs for MA Enrollees by 2018 U.S. Department of Health and Human Services (HHS) silver risk score categories. Figure 2 shows dental and medical PMPM costs for all MA enrollees with dental coverage. As we would expect, allowed medical PMPM costs increase as the risk score increases. However, allowed dental costs PMPM decrease as the risk score increases. Figure 3 shows the allowed dental and medical costs PMPM for people with chronic conditions; again, allowed medical PMPM increases as the risk score increases. Allowed dental PMPM costs follow different patterns depending on the chronic condition. Enrollees with CVD have allowed dental PMPM costs that decrease 60% from the lowest risk score enrollees to the highest risk score enrollees; enrollees with COPD show a slight decline in dental PMPM as the risk scores increase, and enrollees with diabetes and risk scores greater than 15 have dental PMPMs that are half of those with risk scores less than 1.
Note: Enrollees with more than one chronic condition appear more than once in the figures above.
We also combined enrollees with chronic conditions so we could compare their dental and medical allowed PMPM costs to enrollees without chronic conditions. Figure 4 shows dental and medical allowed PMPM costs for enrollees with and without chronic conditions by risk score categories. Enrollees with chronic conditions have dental PMPM costs that decrease more dramatically as risk scores increase than enrollees without chronic conditions. The dental allowed PMPM costs for enrollees with risk scores less than 1 is essentially the same regardless of the presence of chronic conditions.
Note: In the chronic population, an enrollee with more than one chronic condition only appears once.
While we cannot discern the reasons behind the dental PMPM patterns shown in Figures 2 and 3 above, we hypothesize that individuals with chronic conditions may be more consumed with managing their disease, leaving less time and energy to focus on routine dental visits. A lack of awareness of the connections between oral and physical health could contribute to this result as well.
We wonder whether these patterns will be different in the 2021 claim data, as some MAOs begin to offer or encourage use of dental benefits to enrollees with chronic conditions under the MA VBID Model.
Finally, we compared medical and dental allowed costs PMPM for enrollees considered “high utilizers.” We defined high utilizers as enrollees having an annual number of claim lines that exceed the 80th percentile of the total number of medical, dental, and prescription drug claim lines. We again put members into risk score categories. The table in Figure 5 shows that, even within a risk score band indicating similar health status, enrollees who use more medical services also use more dental services. Again, while we cannot ascertain the reasons behind this result, we hypothesize that high utilizers may seek social interaction or comfort from dental as well as medical providers, or may simply seek to get the most out of their benefit plan each year. This result may provide useful insight for predicting dental claims for such individuals. Figure 5 shows the allowed medical and dental PMPM and the ratio of non-high utilizers to high utilizers PMPM costs for MA enrollees.
|Medical PMPM||Dental PMPM|
|Risk Score Bin||Non-HU PMPM||HU PMPM||Ratio of HU to Non-HU||Non-HU PMPM||HU PMPM||Ratio of HU to Non-HU|
|RS Less than 1.0||$218||$1,673||766%||$2.99||$3.81||127%|
|RS Between 1 and 2||$303||$1,433||474%||$2.52||$3.11||123%|
|RS Between 2 and 3||$395||$1,731||438%||$2.86||$3.12||109%|
|RS Between 3 and 5||$448||$1,743||389%||$2.39||$2.84||119%|
|RS Between 5 and 7.5||$557||$2,086||374%||$2.39||$2.60||109%|
|RS Between 7.5 and 10||$786||$2,481||316%||$2.21||$2.47||112%|
|RS Between 10 and 15||$895||$3,031||339%||$1.95||$2.18||112%|
|RS Over 15||$1,769||$6,361||360%||$1.76||$1.67||95%|
The MA VBID Model also encourages MAOs to offer supplemental benefits to enrollees based upon socioeconomic status and to improve health delivery and outcomes to MA enrollees by offering supplemental benefits not typically part of an MA plan and by reducing member cost sharing.6
Supplemental benefits may include, among other benefits, dental service, meal support, and transportation services. Similar to the analyses just described, we explored the utilization patterns and costs of enrollees with social determinants of health (SDOH) diagnoses. Of the 1.9 million MA enrollees with dental coverage in our data, we found almost 16,000 with at least one claim coded with a social determinant of health (SDOH) diagnosis code. Social determinants of health include but are not limited to low educational achievement and employment, housing, and food and family instabilities. As shown in the table in Figure 6, we found little difference in dental utilization for MA enrollees with SDOH diagnoses.
|Dental Services Utilized||Without SDOH||With SDOH|
|No Dental Claims||88%||90%|
|Preventive Dental Claims||11%||9%|
|Other Dental Claims (not preventive)||1%||1%|
The table in Figure 7 shows dental and medical allowed PMPM costs for enrollees with and without SDOH diagnoses. The medical and dental allowed PMPM costs follow the same patterns described earlier: the medical PMPM rises as the risk score rises but the dental PMPM generally falls as the risk score rises. For MA enrollees in the same risk score category, enrollees with SDOH diagnoses have higher medical allowed PMPM cost but dental allowed PMPM costs are similar.
|Medical PMPM||Dental PMPM|
|Risk Score Bin||Non-SDOH||SDOH||Ratio of SDOH to Non-SDOH||Non-SDOH||SDOH||Ratio of SDOH to Non-SDOH|
|RS Less than 1.0||$279||$534||192%||$3.03||$3.11||103%|
|RS Between 1 and 2||$444||$726||164%||$2.59||$2.62||101%|
|RS Between 2 and 3||$610||$939||154%||$2.90||$2.63||91%|
|RS Between 3 and 5||$775||$1,246||161%||$2.51||$2.55||102%|
|RS Between 5 and 7.5||$1,096||$1,604||146%||$2.47||$2.20||89%|
|RS Between 7.5 and 10||$1,565||$2,089||134%||$2.33||$2.37||102%|
|RS Between 10 and 15||$2,030||$2,450||121%||$2.08||$1.63||79%|
|RS Over 15||$5,360||$5,895||110%||$1.69||$1.59||94%|
The Medicare Advantage market provides a unique opportunity to concurrently study dental and medical utilization for a single covered population, and to explore connections between medical and dental cost for people with and without chronic health conditions and by risk score. MAOs participating in the MA VBID Model test have the opportunity to target enrollees with chronic conditions and socioeconomic characteristics with supplemental benefits that could include education on the value of oral health and cost-sharing reductions or other encouragement to obtain dental care as a method to manage these populations. As CMS continues to expect more value and improved health outcomes for its beneficiaries, it will be interesting to watch how dental utilization changes.
We do not know whether the MAOs in our research data have applied to be part of the MA VBID Model nor do we know whether dental benefits will be targeted to enrollees with cardiovascular disease, COPD, diabetes, or social determinants of health. We are unable to determine in our data whether dental benefits are mandatory or optional supplemental benefits nor can we determine which dental services are covered. We understand that MAOs do not use HHS Hierarchical Condition Category (HCC) silver risk scores as their risk adjustment metric. Encounter Data System (EDS) HHS risk scores were not available in our 2018 data but will be calculated in our 2019 research data. SDOH diagnosis codes are not routinely populated. However, out of 1.9 million MA enrollees in our data, we found SDOH diagnosis codes for almost 16,000 enrollees. The lack of a social determinant of health diagnosis does not necessarily mean SDOHs do not exist. ICD-10 diagnosis codes indicating a social determinant of health are those starting with the following three digits: Z55, Z56, Z57, Z58, Z59, Z60, Z62, Z63, Z64, and Z65.
Preventive claims included cleanings, oral exams, and periodontal cleanings, dental procedure codes D1110, D1120, D0120, D0140, D0150, D4341, D4342, and D4910.
We determined enrollees with chronic conditions using HHS-HCC classifications. The table in Figure 8 shows the HHS-HCC classifications we used to determine enrollees with diabetes, cardiovascular disease, and COPD.
|Chronic Conditions||HHS-HCC Classifications, 2018|
|Diabetes||019, 020, 021|
|Cardiovascular Disease||130, 131, 132, 135, 137, 138, 139, 146|
|Enrollees May Appear in More Than One Category|
|Risk Score Category||All||Non-Chronic||Chronic||CVD||COPD||Diabetes|
|RS Less than 1.0||830,731||789,495||41,236||2,004||-||39,348|
|RS Between 1 and 2||246,134||61,552||184,582||874||58,505||138,178|
|RS Between 2 and 3||182,981||130,531||52,450||6,199||5,315||42,120|
|RS Between 3 and 5||179,384||57,914||121,470||45,516||32,928||74,252|
|RS Between 5 and 7.5||142,499||42,677||99,822||63,942||23,277||51,875|
|RS Between 7.5 and 10||76,945||22,071||54,874||38,948||13,716||28,874|
|RS Between 10 and 15||103,303||31,343||71,960||42,646||27,278||36,271|
|RS Over 15||169,198||35,848||133,350||96,857||58,982||71,319|
1CMS. Medicare Advantage Value Based Insurance Design Model. Retrieved November 1, 2020, from https://innovation.cms.gov/innovation-models/vbid.
2Martin, L. Diabetes and Your Smile. ADA. Retrieved November 1, 2020, from https://www.mouthhealthy.org/en/az-topics/d/diabetes.
3Mayo Clinic. Will taking care of my teeth help prevent heart disease? Retrieved November 1, 2020, from https://www.mayoclinic.org/healthy-lifestyle/adult-health/expert-answers/heart-disease-prevention/faq-20057986.
4Gaeckle, N.T. et al. (November 8, 2017). Markers of dental health correlate with daily respiratory symptoms in COPD. Journal of the COPD Foundation. Retrieved November 1, 2020, from https://journal.copdfoundation.org/jcopdf/id/1185/Markers-of-Dental-Health-Correlate-with-Daily-Respiratory-Symptoms-in-COPD.
5Freed, M. et al. (September 18, 2019). Policy Options for Improving Dental Coverage for People on Medicare. Kaiser Family Foundation. Retrieved November 1, 2020, from https://www.kff.org/medicare/issue-brief/policy-options-for-improving-dental-coverage-for-people-on-medicare/.
Understanding dental costs and utilization in the Medicare Advantage population
This paper uses calendar year 2018 Medicare Advantage administrative claim data for enrollees aged 65 and older enrolled in plans with embedded dental coverage to determine the types of dental services used and the dental per member per month costs, as well as to study correlations between medical and dental utilization.