Cost management is not a limitless well, and continued reductions will eventually cut into service capability, quality, or both.
Realizing a level of profitability that can sustain long-term financial viability in the healthcare industry is becoming more difficult as hospitals become pressed for payment reductions by government and major commercial payers. Increases in net patient revenue per equivalent discharge (NPRED) over the past five years have been well below increases in general inflation (1.3 percent compared with 1.7 percent). This has placed increasing efforts on cost management and improved efficiency to maintain profit margins.
Cost management is not a limitless well, and continued reductions eventually will cut into service capability, quality, or both. Various factors can create payment variation, and healthcare leaders can isolate areas that maintain or improve performance.
Sources of Net Patient Revenue
Patient revenue is derived from three major sources in the healthcare industry; government payers, commercial payers, and uninsured patients.
Governmental payers. Creating positive changes in net revenues from government programs, largely Medicare and Medicaid, has become increasingly difficult because federal and state governments are experiencing budgetary problems and are looking for areas to reduce their expenditures.
Most likely, changes in eligibility, benefit structure, and actual service payment levels will be made in the next decade. Risk will be shifted to providers wherever feasible. Healthcare providers can turn to various methods to increase payment.
First, accurate coding of claims can have a positive impact on payment. For example, identifying existing complications in the medical record can produce a DRG shift with potentially higher payment.
Second, service mix is also a factor. For example, surgical procedures have historically received higher payment relative to medical procedures, and behavioral care is often subject to lower payment levels.
Finally, higher volume can produce greater revenue, but not necessarily greater revenue per equivalent unit of service, unless the additional volume is in areas where pricing is more favorable (e.g., a shift in payer mix to more commercial business).
Commercial payers. Health plans can be split into two categories; those with contracts and those without. For health plans with provider contracts, the most significant driver of increased revenue is contract negotiation. Payment terms in managed care contracts vary widely across providers in the same region. Some difference may be attributable to either market share or product differentiation, but a large portion is simply related to skillful negotiation.
Other factors can also improve net revenues given existing contract terms such as coding and pricing. We have seen with increasing frequency revenue being lost because prices are lower than fee-schedule amounts. For health plans without contracts, pricing is a critical factor. Collection effort can also be an issue with many commercial health plans as they try to reduce payment by arguing for reasonableness of charges or medical necessity.
Uninsured patients. This category may become larger in the future for several reasons.
First, the future of the Affordable Care Act could have a sizable influence on the size of the uninsured population in the United States.
Second, the level of patient responsibility is increasing as employers and payers seek to shift more of the payment and risk to the beneficiary. Insured plans with deductibles of $5,000 are increasingly common and increases in coinsurance levels are also rising. Billing and collection activities have become the keys to improving collection experience in this arena.
Assessing Payment Differences Across Hospitals
We believe the best way to assess payment differences across hospitals is a comparison of actual payments made for like services across all payers.
For example, assume that there is only one patient encounter type—a simple pneumonia case. Using a hypothetical illustration of a two-hospital payer comparison, both hospitals have received an average payment of $5,000 per case, but Hospital A has received $14,000 per private payer case, while Hospital B has received $5,111 ($46,000/9). Both hospitals receive $4,000 per case from the government, which requires Hospital A to negotiate a much larger private payer payment because 90 percent of its business is governmental compared with 10 percent at Hospital B.
This pattern of payment illustrates “cost shifting” in the healthcare marketplace. Hospitals and other healthcare providers with high levels of governmental patients, especially Medicaid, and high levels of medically indigent patients must receive larger payments from commercial health plans to remain solvent. Commercial health plans may argue they should not be held responsible for poor payer mixes, but the reality is someone must provide higher payment levels to providers who are receiving inadequate payments from the government and so indigent patients can remain financially viable.
Impact of Ownership
Our focus for the remainder of this article is directed at the hospital sector largely because of the presence of available data and also the relative size of the sector in the healthcare industry. Specifically, we will examine key relationships related to NPRED, which measures the realization of revenue for all inpatients and outpatients across all payers. The metric is defined as net patient revenue, the value reported in Worksheet G of the Medicare Cost Report, divided by equivalent discharges. Equivalent discharge is a metric that we believe is more valid than adjusted discharges as a measure of facility output (Cleverley, W., “ Time to Replace Adjusted Discharges”, hfm, Healthcare Financial Management Association, May 2014).
We computed NPRED for all prospectively paid hospitals in the United States for 2015. Investor-owned hospitals have the highest level of payment revenue realized per equivalent unit of service followed closely by governmental hospitals. Operating margins are also highest for investor-owned hospitals. Our focus will be directed to voluntary not-for-profit (VNP) hospitals because we believe that the governing boards of voluntary hospitals restrict both the level of payment that can be realized and the amount of profit necessary to maintain financial solvency.
Data and Impact Factors
Using the VNP data base for 2015, we next removed hospitals with less than $100 million in 2015 net patient revenue. This left 1,270 VNP hospitals in the study. These hospitals accounted for 88 percent of the total 2015 net patient revenue of all VNP hospitals and 63 percent of the total net patient revenue for all prospectively paid hospitals. The following factors were included in the study based upon prior identification as determinants of either price or profitability.
Surgery percent. Surgery cases have been viewed as more profitable than medical cases because of the perception of higher net revenue to cost relationships. We have defined surgery percent as the ratio of 2015 inpatient Medicare surgery discharges to total inpatient Medicare discharges.
Inpatient revenue percent. Historically, inpatient business has been viewed as less profitable than outpatient revenue but Medicare margins have historically been higher for inpatient business relative to outpatient business. We have defined inpatient revenue percent as the ratio of total inpatient charges to total charges from Worksheet G-2 in the Medicare Cost Report.
Medicare case mix index (CMI). The common belief is that a higher CMI will increase prices but using the equivalent discharge metric adjusts for both inpatient and outpatient complexities so the actual relationship projected is not clear.
Complication and comorbidity capture percent. The expectation is that the presence of complications or comorbidities (CC) will increase prices but the equivalent discharge metric adjusts for much of the variance assuming cases with a CC or major complication or comorbidity (MCC) have higher case mix payment rates. CC capture percent is defined as Medicare inpatient cases with a complication or comorbidity divided by the total number of Medicare inpatient cases where a complication or comorbidity designation is present.
Hospital charge index (HCI). The HCI is a metric that provides a relative assessment of the charges made to both inpatient and outpatient encounters adjusting for both case mix and cost of living differences. The expectation is that higher charges should lead to higher net revenue collection. (Cleverley, W., “Effective Hospital Pricing Strategy,” hfm, Healthcare Financial Management Association, April 2003).
Teaching status. Teaching hospitals are perceived as being higher priced because of a greater case complexity but adjustments for both inpatient and outpatient case mix present in the equivalent discharge metric should remove much of the difference.
Hospital quality index (HQI). The HQI is a weighted metric using publicly available measures from the Hospital Compare website. National mortality metrics receive 40 percent of the weight, national readmission metrics receive 20 percent of the weight, national complication metrics receive 10 percent of the weight, and the remaining 30 percent is assigned to the process of care metrics. While everyone would agree about the importance of quality, its impact on actual revenue realization is unclear. (Cleverley, W., and Cleverley, J. “Is There a Cost Associated with Higher Quality,” hfm, Healthcare Financial Management Association, January 2011).
Market share. In general, the belief is that larger market share should enable better managed care contract terms through enhanced market leverage. We have defined market share as the ratio of net patient revenue for the hospital to the sum of all hospital net patient revenue in the county in which the hospital is located.
Cost per equivalent discharge (CPED) wage index adjusted. CPED should be the key driver of NPRED because hospitals must recover their costs and there is an expectation that the board of a VNP hospital would limit the realization of excessive levels of profitability. CPED is defined as total payable expenses from Worksheet A of the Medicare Cost Report divided by equivalent discharge.
Uncompensated care cost (UCC). UCC is the total amount of uncompensated care cost from Worksheet S-10 of the Medicare Cost Report divided by total operating expenses. UCC includes the difference between Medicaid payments and Medicaid costs plus charity care. It does not include unpaid Medicare costs. The expectation is that higher UCC levels reduce NPRED.
Equivalent discharges. Equivalent discharge is a measure of scale of operations. Economic theory would suggest that larger scale should lower cost per unit and as a result lower NPRED assuming VNP hospital boards restrict profitability.
Using the variables identified above, we ran a multiple regression with NPRED as the dependent variable with the 1,270 VNP hospitals in the data set. The mean NPRED was $8,932, and the median was $8,732. Of the 11 variables in the regression, 10 of them were statistically significant at the 2 percent level or lower. Only the size measure, equivalent discharge, was not significant.
To provide some relative assessment of the importance of each variable, we used the regression coefficient and multiplied that value times the mean or average value of the variable times 10 percent. This would provide a measure of the possible impact of a 10 percent increase in the variable. For example, the mean value of surgical cases was 27.46 percent, and a 10 percent increase in that value would be 2.746 percent. Multiplying that value by the regression coefficient yields a value of $112.73 (2.746 percent x 41.04). We can then use this value to compare with other variables (e.g., surgery changes appear to be much more influential on NPRED than changes in uncompensated care [-17.89]).
The impact of CPED upon NPRED was by far the largest force driving NPRED—a 10 percent increase in CPED of $873.60 would be expected to increase NPRED by $670.69. VNP hospitals need to set NPRED to levels that will cover their costs as do most other hospitals, but what was surprising was the significance and impact of other factors. Nine of the remaining 10 metrics were statistically significant.
A 10 percent increase in CMI would reduce NPRED by $166.80. This may seem illogical, but the equivalent discharge metric is case adjusted for both inpatient and outpatient and the metric used here is inpatient only. There is also a significant correlation between surgery percentages and CMI. Considering the effects of both variables suggests that increases in case mix that are associated with medical cases as opposed to surgery cases are likely to reduce NPRED because the metric is already adjusted for case mix. In short, case weights associated with medical cases appear to be underweighted relative to surgery cases. This is consistent with Medicare payment changes that have tried to shift relative payment weights toward medical cases.
The HQI was statistically significant and had a positive coefficient, which suggests that hospitals with higher quality scores do in fact receive some financial adjustments to their net revenue. This increase may be associated with the hospital’s ability to differentiate its services from others and leverage that position into higher commercial contract payments. There was a small negative correlation between CPED and HQI, but the relationship was not significant. This suggests that higher cost is not associated with better quality.
Market share was also a statistically significant factor that influenced NPRED, but it did not have a sizable influence—a 10 percent increase in relative market share would add only $16.72 to NPRED. Some may believe the inclusion of equivalent discharge in the regression equation might dilute the impact of market share, but we found that there was no significant correlation between equivalent discharge and market share. We did note that there was a negative correlation between wage index and market share which suggests the higher market share values were more likely to be in lower cost of living areas which might be less urban. In some of the larger urban areas, it is also important to note that we did not aggregate market share for hospitals that were part of the same system. This might have also weakened the relationship.
Complication and comorbidity had a positive and sizable impact on NPRED, a 10 percent increase in the value would add $96.58 to NPRED. This suggests that as the percentage of DRGs that have a CC and MCC alternative increase, the level of NPRED increases more than the effect produced by the increase in CMI. Improved recognition of complications and comorbidities during the coding process appears to result in significant improvements in revenue.
Pricing, which was represented by the HCI, had a positive and significant effect on NPRED. A 10 percent increase in relative pricing would increase NPRED by $82.10. Increases in prices produce higher net revenues for contract areas that are paid on percent-of-charge bases or cases in which outliers or lesser-than provisions exist.
Increases in uncompensated care have a small but significant influence on NPRED. Increases in uncompensated care would produce a need to cost shift to other payers. The negative value suggests that full cost shifting is not always possible.
Teaching hospitals have lower NPRED than non-teaching hospitals, but the impact is very small. There is a high correlation between teaching status and CPED, which might mean that separation of the effect between the two might be difficult.
Finally, hospitals with larger percentages of inpatient revenue appear to realize slightly lower levels of NPRED. This may suggest there is a greater ability to realize higher margins on outpatient procedures relative to inpatient procedures. Historically, this may have been true because hospital outpatient procedures represented a much smaller percent of commercial payers’ total book of business, but this position has been changing quickly, and hospitals are beginning to realize they may have more market leverage in inpatient areas as compared with outpatient procedures, where a substantial number of alternative providers exist.
Some of the unexpected findings in our research include the following conclusions:
- Patient service mix was an important factor—increases in inpatient case mix exert a negative influence on NPRED unless the increase in case mix is concentrated in surgical areas.
- Market share has a relatively small influence on NPRED.
- Hospital quality improvement does lead to higher levels of NPRED and surpasses the influence of market share.
- Teaching hospitals do not have higher levels of NPRED, which suggests that when payments are adjusted for both inpatient and outpatient, there is no advantage for teaching hospitals.
Healthcare finance leaders can have a positive impact on payment by reviewing variables and targeting those that help balance payers that provide lower payments.