TheJournalofEmergency
Medicine,
Vol 10, pp 401-405,
Printed in the USA . Copyright 0 1992 Pergamon PressLtd.
1992
PEER REVIEW ORGANIZATION PAYMENT DENIALS: COMPARATIVE ANALYSIS OF EMERGENCY DEPARTMENT AND NON-EMERGENCY-DEPARTMENT ADMISSIONS David T. Overton, MD, MBA, Penny Sokolowski,
FACEP,*t RN,
BSN,§
Michael S. Kobernick, MD, FACEP,t* and Valerie Belcher, RN, BSN$
‘Department of Emergency Medicine, Michigan State University, Kalamazoo Center for Medical Studies, Kalamazoo, Michigan; TSection of Emergency Medicine, Department of Surgery, University of Michigan Medical Center, Ann Arbor, Michigan; *Department of Emergency Medicine, Macomb Hospital Center, Warren, Michigan; QMedical Quality Program Management, William Beaumont Hospital, Royal Oak, Michigan Reprint Address: David T. Overton, MD, FACEP, Department of Emergency Medicine, Kalamazoo Center for Medical Studies, 1535 Gull Road, Suite 230, Kalamazoo, Michigan 49001
0 Abstract-The Health Care Financing Administration has contracted with regional peer review organizations to review Medicare admissions and to deny payment for hospital admissions that fail to meet peer review organization criteria. The purpose of this study was to compare emergency department admissions with non-emergency-department admissions with respect to rates of peer review organization denial and the reasons for those denials. All hospital Medicare admissions between January 1984 and April 1987 were retrospectively reviewed. Patients were excluded if they received peer review organization preauthorization prior to admission. The rest were classified by 1) source of admission (emergency department or nonemergency department), 2) peer review organization decision, 3) reason for peer review organization denial, 4) whether the denial was appealed, 5) the results of appeal. Chi-square or Fisher’s Exact Test analysis was performed, and P < 0.05 was considered to be significant. During the 40-month study period, there were 19,847 emergency department Medicare admissions and 19,752 non-emergency-department Medicare admissions. Of the nonemergency-department admissions, 7887 received pre-authorization. None of the emergency department admissions received pre-authorfzatfon. Of the 19,847 emergency department admissions, 433 (2.23@70)were denied. Of these denials, 269 (60.77%) were appealed by the hospital; 136 (50.5%) successfufly. Of the 11,865 non-emergency department, non-pre-authorized admissions, 333 (2.81%) were
denied. Of these denials, 174 (52.29’0) were appealed, 76 (43.6%) successfully. Overall, emergency department admissions were significantly lesslikely to receive peer review organization denial than non-emergency-department, nonpre-authorized admissions (P < 0.003). When we looked at individual categories for denial, only Severity of Illness and Fragmented Care had significant differences (P < 0.00002, and P < 0.002, respectively). We conclude that, in this setting, emergency department Medicare admfssions are at lower risk of peer review organization Medicare denial than non-emergency-department, non-pre-authorized Medicare admissions. This difference is due to decreased denials in the Severity of fIbtess and Fragmented Care categories. 0 Keywords-peer review; professional review organizations; utilization review; emergency services,hospital
INTRODUCTION
The Tax Equity and Fiscal ResponsibilityAct (TEFRA) of 1982(PL 97-248)directedthe Health Care FinancingAdministration (HCFA) to reviewhospital Medicareadmissionsnationwidefor appropriateness and necessityof medicalcarerendered(1,2). Section 141of TEFRA requiresHCFA to contractwith physician-sponsored or physician-access organizationsto meetthis responsibility. These peer review organiza-
Presented at the American College of Emergency Physicians Scientific Assembly, September, 1988, New Orleans.
Ortginai Contributions presents articles of interest to both academic and practicing physicians. This section of JEM is coordinated by John A. MWX, MD, of Carolinas Medical Center, Charlotte, North Carolina. RECEIVED: 28 May 1991;FINAL SUBMISSION RECEIVED: 7 November1991; 0736-4679/92$5.00+ .OO =
ACCEPTED:
2 December1991
401
402
D. T. Overton,
tions (PRO) evaluate Medicare admissions and deny payment to hospitals for admissions failing to meet PRO criteria. In addition, legislation prohibits hospitals from directly billing Medicare beneficiaries for any payments denied by the PRO. Thus, PRO payment denials are usually unrecoverable losses to hospitals. In many hospitals, these denials have resulted in substantial financial losses (3). Prior studies have shown that Medicare admissions originating in the emergency department (ED) are significantly more costly to hospitals than nonED admissions, and may represent a financial liability to institutions under the DRG system (4-6). There is an impression among some utilization review personnel that ED admissions represent a disproportionate number of these PRO denials. To our knowledge, this impression is unsubstantiated in the medical literature. Thus, our hypothesis was that a higher percentage of ED Medicare admissions are rejected than non-ED Medicare admissions. We specifically examined ED admissions and non-ED admissions, comparing the reasons for PRO denial and rates of denial.
METHODS William Beaumont Hospital is a 940-bed community teaching hospital. There are about 72,000 ED patient registrations annually. Approximately 22% of this population are admitted to the hospital, and approximately 17% of the ED patients seen are Medicare beneficiaries. The majority of hospitalized patients are cared for by housestaff in conjunction with attending physicians. All Medicare admissions between January 1984 (when TEFRA started review) and April 1987 were retrospectively reviewed. Excluded were patients receiving PRO preauthorization prior to admission. Certain Medicare patients are required to receive such pre-authorization (7), usually for elective inpatient surgical procedures. Physicians must contact the PRO in advance and supply information regarding the patient. The PRO then decides whether the patient meets criteria for admission. (The in-hospital care of such patients is reviewed by the PRO, and subsequent denial may still be issued in some instances.) Pre-authorized patients were excluded from our study, as they are not comparable to the non-prescreened population of admissions, in which PRO decisions are made after hospital admission takes place. All other admissions are classified according to:
M. S. Kobernick,
P. Sokolowski,
V. Belcher
1) source of admission: ED or non-ED, 2) PRO decision, 3) reason for PRO denial, 4) whether the denial was appealed, and 5) the results of the appeal. There are a variety of reasons why the PRO may deny payment to a hospital for a Medicare patient admission: Severity of Illness-The patient did not have sufficient Severity of Illness to warrant hospitalization, and treatment could have been carried out on an outpatient basis. These decisions are based on objective criteria, with the final review decision being made by a physician (8). Intensity of Service-Regardless of the Severity of Illness, a patient must receive a sufficient Intensity of Service while in the hospital. This generally means care that could not have been delivered as an outpatient. This, too, is based on specific criteria. DRG - The PRO may disagree with the principal diagnosis DRG that the hospital submits, preferring another principal diagnosis (7). Day Outlier-This is a denial of the hospital’s claim to reimbursement under the day outlier system (7). Cost Outlier-This is a denial of the hospital’s claim to reimbursement under the cost outlier system (7). Premature Discharge - If patients are readmitted soon after discharge, the PRO may maintain that they were discharged prematurely the first time and deny payment for the second admission. Fragmented Care-Two clearly related admissions (that is, a diagnostic biopsy followed by surgery) often should be submitted as one bill. If they are not, the PRO may deny payment until they are correctly submitted. Technical Chart Error-This represents an inability of the hospital to produce a chart for the PRO to review. Technical Billing Error-This represents a clerical error in the preparation of the bill. Finally, many hospitals have set up internal review systems to detect, as soon as possible after admission, patients at high risk for PRO denial. Hospitals often do not even submit a bill to Medicare for these cases because there are incentives for hospitals to keep their total percentage of denials as low as possible. In our hospital, this is termed an In-House Denial and represents a variation of a Severity of Illness deficiency.
403
PRO Payment Denials
Regardless of the reason for the initial notice of denial, hospitals are entitled to appeal the PRO decision by supplying additional documentation. Our hospital review and appeal process is similar for both ED and non-ED admissions. Examination of the above reasons reveals that Severity of Illness (and thus, In-house Denial) is directly under the control of the emergency physician making the decision to admit a Medicare patient to the hospital. It might also be reasoned that some factors influencing Intensity of Service, such as cardiac monitoring or IV antibiotics, might be under emergency physician influence. All other factors, however, reflect care rendered or decisions made by other physicians or hospital personnel after the patient is admitted.
StatisticalAnalysis ED admissions and non-ED, non-pre-authorized admissions were compared, using either chi-square analysis or Fisher’s Exact Test, as appropriate. A P level of less than < 0.05 was considered significant.
StatisticalAnalysis Statistical analysis comparing ED admissions to nonED admissions is shown in Table 3, If Severity of Illness denials alone are considered, emergency department admissions were statistically less likely to receive PRO denial than nonemergency-department admissions (P < 0.00002).If Fragmented Care denials alone are considered, emergency department admissions were also less likely to receive denials (P < 0.002). When all reasons for denial are considered, significantly fewer emergency department admissions received denials than did non-emergency-department admissions (P < 0.003). The strongly significant Severity of Illness category along with the Fragmented Care category influenced these total figures. If these denials are excluded, the Total comparisons lose their significance. For reasons of PRO denial other than Severity of Illness and Fragmented Care, no statistically significant differences were detected.
DISCUSSION RESULTS During the 40-month study period, there were a total of 133,790 admissions to William Beaumont Hospital; 39,599 (29.6%) were Medicare admissions. Of the Medicare admissions, 19,847 (50.12%) were through the ED, and 19,752 (49.88%) were non-ED. PRO pre-authorization was obtained in 7887 (39.93%) of the non-ED Medicare admissions. None of the ED Medicare admissions received preauthorization for admission.
EmergencyDepartmentAdmissions Of the 19,847 emergency department Medicare admissions, 443 (2.23%) received PRO denials (Table 1). Of these 443 denials, 269 (60.7%) were appealed; 137 (50.9%) were successful.
Non-Emergency-Department Admissions Of the 11,865 non-ED, non-pre-authorized Medicare admissions, 333 (2.816/o) were denied (Table 2). Of the 333 denials, 174 (52.2%) were appealed; 77 (44.2%) were successful.
This study reveals a lower PRO denial rate for ED admissions. However, a number of other factors may affect hospital financial impact. Prior studies have shown that Medicare admissions originating in the ED are significantly more costly to hospitals than non-ED admissions and represent a financial liability to institutions (4-6). This has been shown for Medicare admissions in general (4-6) as well as for specific categories, such as surgical (5) and neurosurgical admissions (9). It has been claimed that the DRG system does not adequately compensate institutions for intensive care treatment of critically injured patients (lo), and these authors maintain that a DRG modifier is needed to correct this situation. Other authors have pointed out that clinical judgment outperforms objective PRO Severity of Illness criteria in identifying patients who need acute admission (11). In contrast, we have shown that in our patient population and practice setting, emergency department admissions are at decreased risk for PRO denial and do not represent a disproportionate financial liability to our hospital with regard to such denials. The differences in the Severity of Illness and Fragmented Care categories are statistically significant and account for these differences. It is not surprising that denials for Fragmented Care might be more common in non-emergency-
404
D. T. Overton, Table 1. Emergency Department Medicare Admission (Total Admissions-19,847) Reason for Denial
M. S. Kobernick,
P. Sokolowski,
Denials
Didn’t appeal
Appeal denied
Appeal won
Total
33 9 8 109 8 6 1 0 0 0 174
48 15 NA 61 4 1 3 0 0 0 132
66 5 NA 39 11
147 29 8 209 23 12 12 0 2 1 443
Severity of Illness Intensity of Service In-House Denial DRG Day Outlier Cost Outlier Premature Discharge Fragmented Care Technical-Chart Technical-Bill Totals
ii 0 2 1 137
NA = not applicable.
Table 2. Non-Emergency-Department (Total Admlsslons-11,888) Reason for Denial
Medicare Admission
Didn’t appeal
Severity of Illness Intensity of Service In-House Denial DRG Day Outlier Cost Outlier Premature Discharge Fragmented Care Technical-Chart Technical-Bill Totals
Denials
Appeal denied
Appeal won
Total
47 3 NA 38 6 1 2 0 0 0 97
45 6 NA 14 6 0 1 4 1 0 77
146 16 10 125 21 4 3 7 1 0 333
54 7 10 73 9 3 0 z 0 159
NA = not applicable.
Table 3. Comparison Admlsslons
between Emergency
Reason for Denial Severity of Illness Intensity of Service In-House Denial DRG Day Outlier Cost Outlier Premature Discharge Fragmented Care Technical-Chart Technical-Bill Severity of Illness and In-House Denial Totals NA = not applicable.
Didn’t appeal P
Medicine,
Vol 10, pp 401-405,
Printed in the USA . Copyright 0 1992 Pergamon PressLtd.
1992
PEER REVIEW ORGANIZATION PAYMENT DENIALS: COMPARATIVE ANALYSIS OF EMERGENCY DEPARTMENT AND NON-EMERGENCY-DEPARTMENT ADMISSIONS David T. Overton, MD, MBA, Penny Sokolowski,
FACEP,*t RN,
BSN,§
Michael S. Kobernick, MD, FACEP,t* and Valerie Belcher, RN, BSN$
‘Department of Emergency Medicine, Michigan State University, Kalamazoo Center for Medical Studies, Kalamazoo, Michigan; TSection of Emergency Medicine, Department of Surgery, University of Michigan Medical Center, Ann Arbor, Michigan; *Department of Emergency Medicine, Macomb Hospital Center, Warren, Michigan; QMedical Quality Program Management, William Beaumont Hospital, Royal Oak, Michigan Reprint Address: David T. Overton, MD, FACEP, Department of Emergency Medicine, Kalamazoo Center for Medical Studies, 1535 Gull Road, Suite 230, Kalamazoo, Michigan 49001
0 Abstract-The Health Care Financing Administration has contracted with regional peer review organizations to review Medicare admissions and to deny payment for hospital admissions that fail to meet peer review organization criteria. The purpose of this study was to compare emergency department admissions with non-emergency-department admissions with respect to rates of peer review organization denial and the reasons for those denials. All hospital Medicare admissions between January 1984 and April 1987 were retrospectively reviewed. Patients were excluded if they received peer review organization preauthorization prior to admission. The rest were classified by 1) source of admission (emergency department or nonemergency department), 2) peer review organization decision, 3) reason for peer review organization denial, 4) whether the denial was appealed, 5) the results of appeal. Chi-square or Fisher’s Exact Test analysis was performed, and P < 0.05 was considered to be significant. During the 40-month study period, there were 19,847 emergency department Medicare admissions and 19,752 non-emergency-department Medicare admissions. Of the nonemergency-department admissions, 7887 received pre-authorization. None of the emergency department admissions received pre-authorfzatfon. Of the 19,847 emergency department admissions, 433 (2.23@70)were denied. Of these denials, 269 (60.77%) were appealed by the hospital; 136 (50.5%) successfufly. Of the 11,865 non-emergency department, non-pre-authorized admissions, 333 (2.81%) were
denied. Of these denials, 174 (52.29’0) were appealed, 76 (43.6%) successfully. Overall, emergency department admissions were significantly lesslikely to receive peer review organization denial than non-emergency-department, nonpre-authorized admissions (P < 0.003). When we looked at individual categories for denial, only Severity of Illness and Fragmented Care had significant differences (P < 0.00002, and P < 0.002, respectively). We conclude that, in this setting, emergency department Medicare admfssions are at lower risk of peer review organization Medicare denial than non-emergency-department, non-pre-authorized Medicare admissions. This difference is due to decreased denials in the Severity of fIbtess and Fragmented Care categories. 0 Keywords-peer review; professional review organizations; utilization review; emergency services,hospital
INTRODUCTION
The Tax Equity and Fiscal ResponsibilityAct (TEFRA) of 1982(PL 97-248)directedthe Health Care FinancingAdministration (HCFA) to reviewhospital Medicareadmissionsnationwidefor appropriateness and necessityof medicalcarerendered(1,2). Section 141of TEFRA requiresHCFA to contractwith physician-sponsored or physician-access organizationsto meetthis responsibility. These peer review organiza-
Presented at the American College of Emergency Physicians Scientific Assembly, September, 1988, New Orleans.
Ortginai Contributions presents articles of interest to both academic and practicing physicians. This section of JEM is coordinated by John A. MWX, MD, of Carolinas Medical Center, Charlotte, North Carolina. RECEIVED: 28 May 1991;FINAL SUBMISSION RECEIVED: 7 November1991; 0736-4679/92$5.00+ .OO =
ACCEPTED:
2 December1991
401
402
D. T. Overton,
tions (PRO) evaluate Medicare admissions and deny payment to hospitals for admissions failing to meet PRO criteria. In addition, legislation prohibits hospitals from directly billing Medicare beneficiaries for any payments denied by the PRO. Thus, PRO payment denials are usually unrecoverable losses to hospitals. In many hospitals, these denials have resulted in substantial financial losses (3). Prior studies have shown that Medicare admissions originating in the emergency department (ED) are significantly more costly to hospitals than nonED admissions, and may represent a financial liability to institutions under the DRG system (4-6). There is an impression among some utilization review personnel that ED admissions represent a disproportionate number of these PRO denials. To our knowledge, this impression is unsubstantiated in the medical literature. Thus, our hypothesis was that a higher percentage of ED Medicare admissions are rejected than non-ED Medicare admissions. We specifically examined ED admissions and non-ED admissions, comparing the reasons for PRO denial and rates of denial.
METHODS William Beaumont Hospital is a 940-bed community teaching hospital. There are about 72,000 ED patient registrations annually. Approximately 22% of this population are admitted to the hospital, and approximately 17% of the ED patients seen are Medicare beneficiaries. The majority of hospitalized patients are cared for by housestaff in conjunction with attending physicians. All Medicare admissions between January 1984 (when TEFRA started review) and April 1987 were retrospectively reviewed. Excluded were patients receiving PRO preauthorization prior to admission. Certain Medicare patients are required to receive such pre-authorization (7), usually for elective inpatient surgical procedures. Physicians must contact the PRO in advance and supply information regarding the patient. The PRO then decides whether the patient meets criteria for admission. (The in-hospital care of such patients is reviewed by the PRO, and subsequent denial may still be issued in some instances.) Pre-authorized patients were excluded from our study, as they are not comparable to the non-prescreened population of admissions, in which PRO decisions are made after hospital admission takes place. All other admissions are classified according to:
M. S. Kobernick,
P. Sokolowski,
V. Belcher
1) source of admission: ED or non-ED, 2) PRO decision, 3) reason for PRO denial, 4) whether the denial was appealed, and 5) the results of the appeal. There are a variety of reasons why the PRO may deny payment to a hospital for a Medicare patient admission: Severity of Illness-The patient did not have sufficient Severity of Illness to warrant hospitalization, and treatment could have been carried out on an outpatient basis. These decisions are based on objective criteria, with the final review decision being made by a physician (8). Intensity of Service-Regardless of the Severity of Illness, a patient must receive a sufficient Intensity of Service while in the hospital. This generally means care that could not have been delivered as an outpatient. This, too, is based on specific criteria. DRG - The PRO may disagree with the principal diagnosis DRG that the hospital submits, preferring another principal diagnosis (7). Day Outlier-This is a denial of the hospital’s claim to reimbursement under the day outlier system (7). Cost Outlier-This is a denial of the hospital’s claim to reimbursement under the cost outlier system (7). Premature Discharge - If patients are readmitted soon after discharge, the PRO may maintain that they were discharged prematurely the first time and deny payment for the second admission. Fragmented Care-Two clearly related admissions (that is, a diagnostic biopsy followed by surgery) often should be submitted as one bill. If they are not, the PRO may deny payment until they are correctly submitted. Technical Chart Error-This represents an inability of the hospital to produce a chart for the PRO to review. Technical Billing Error-This represents a clerical error in the preparation of the bill. Finally, many hospitals have set up internal review systems to detect, as soon as possible after admission, patients at high risk for PRO denial. Hospitals often do not even submit a bill to Medicare for these cases because there are incentives for hospitals to keep their total percentage of denials as low as possible. In our hospital, this is termed an In-House Denial and represents a variation of a Severity of Illness deficiency.
403
PRO Payment Denials
Regardless of the reason for the initial notice of denial, hospitals are entitled to appeal the PRO decision by supplying additional documentation. Our hospital review and appeal process is similar for both ED and non-ED admissions. Examination of the above reasons reveals that Severity of Illness (and thus, In-house Denial) is directly under the control of the emergency physician making the decision to admit a Medicare patient to the hospital. It might also be reasoned that some factors influencing Intensity of Service, such as cardiac monitoring or IV antibiotics, might be under emergency physician influence. All other factors, however, reflect care rendered or decisions made by other physicians or hospital personnel after the patient is admitted.
StatisticalAnalysis ED admissions and non-ED, non-pre-authorized admissions were compared, using either chi-square analysis or Fisher’s Exact Test, as appropriate. A P level of less than < 0.05 was considered significant.
StatisticalAnalysis Statistical analysis comparing ED admissions to nonED admissions is shown in Table 3, If Severity of Illness denials alone are considered, emergency department admissions were statistically less likely to receive PRO denial than nonemergency-department admissions (P < 0.00002).If Fragmented Care denials alone are considered, emergency department admissions were also less likely to receive denials (P < 0.002). When all reasons for denial are considered, significantly fewer emergency department admissions received denials than did non-emergency-department admissions (P < 0.003). The strongly significant Severity of Illness category along with the Fragmented Care category influenced these total figures. If these denials are excluded, the Total comparisons lose their significance. For reasons of PRO denial other than Severity of Illness and Fragmented Care, no statistically significant differences were detected.
DISCUSSION RESULTS During the 40-month study period, there were a total of 133,790 admissions to William Beaumont Hospital; 39,599 (29.6%) were Medicare admissions. Of the Medicare admissions, 19,847 (50.12%) were through the ED, and 19,752 (49.88%) were non-ED. PRO pre-authorization was obtained in 7887 (39.93%) of the non-ED Medicare admissions. None of the ED Medicare admissions received preauthorization for admission.
EmergencyDepartmentAdmissions Of the 19,847 emergency department Medicare admissions, 443 (2.23%) received PRO denials (Table 1). Of these 443 denials, 269 (60.7%) were appealed; 137 (50.9%) were successful.
Non-Emergency-Department Admissions Of the 11,865 non-ED, non-pre-authorized Medicare admissions, 333 (2.816/o) were denied (Table 2). Of the 333 denials, 174 (52.2%) were appealed; 77 (44.2%) were successful.
This study reveals a lower PRO denial rate for ED admissions. However, a number of other factors may affect hospital financial impact. Prior studies have shown that Medicare admissions originating in the ED are significantly more costly to hospitals than non-ED admissions and represent a financial liability to institutions (4-6). This has been shown for Medicare admissions in general (4-6) as well as for specific categories, such as surgical (5) and neurosurgical admissions (9). It has been claimed that the DRG system does not adequately compensate institutions for intensive care treatment of critically injured patients (lo), and these authors maintain that a DRG modifier is needed to correct this situation. Other authors have pointed out that clinical judgment outperforms objective PRO Severity of Illness criteria in identifying patients who need acute admission (11). In contrast, we have shown that in our patient population and practice setting, emergency department admissions are at decreased risk for PRO denial and do not represent a disproportionate financial liability to our hospital with regard to such denials. The differences in the Severity of Illness and Fragmented Care categories are statistically significant and account for these differences. It is not surprising that denials for Fragmented Care might be more common in non-emergency-
404
D. T. Overton, Table 1. Emergency Department Medicare Admission (Total Admissions-19,847) Reason for Denial
M. S. Kobernick,
P. Sokolowski,
Denials
Didn’t appeal
Appeal denied
Appeal won
Total
33 9 8 109 8 6 1 0 0 0 174
48 15 NA 61 4 1 3 0 0 0 132
66 5 NA 39 11
147 29 8 209 23 12 12 0 2 1 443
Severity of Illness Intensity of Service In-House Denial DRG Day Outlier Cost Outlier Premature Discharge Fragmented Care Technical-Chart Technical-Bill Totals
ii 0 2 1 137
NA = not applicable.
Table 2. Non-Emergency-Department (Total Admlsslons-11,888) Reason for Denial
Medicare Admission
Didn’t appeal
Severity of Illness Intensity of Service In-House Denial DRG Day Outlier Cost Outlier Premature Discharge Fragmented Care Technical-Chart Technical-Bill Totals
Denials
Appeal denied
Appeal won
Total
47 3 NA 38 6 1 2 0 0 0 97
45 6 NA 14 6 0 1 4 1 0 77
146 16 10 125 21 4 3 7 1 0 333
54 7 10 73 9 3 0 z 0 159
NA = not applicable.
Table 3. Comparison Admlsslons
between Emergency
Reason for Denial Severity of Illness Intensity of Service In-House Denial DRG Day Outlier Cost Outlier Premature Discharge Fragmented Care Technical-Chart Technical-Bill Severity of Illness and In-House Denial Totals NA = not applicable.
Didn’t appeal P
