1 patterns of quality for patients with three chronic conditions stephen m. davidson, ph.d. michael...
TRANSCRIPT
1
Patterns of Quality for Patients with Three Chronic Conditions
Stephen M. Davidson, Ph.D.
Michael Shwartz, Ph.D.
Randall S.Stafford, M.D.
Academy of Health Services Research and Health Policy
San Diego, CA
June 7, 2004
2
It is widely accepted that quality of care is suboptimal and unreliable– Wide variation in HEDIS measures– Deficits in condition-specific patterns of care– IOM reports– Large numbers of published stories of errors
3
Although uncertain safety and quality of health care are serious and pervasive problems, the actual extent to which inadequate care is found in the U.S. is unknown.
One reason: difficulty of measurement.
4
Three Main Goals
1. Describe new claims-based measures showing 5 levels of quality for each of 3 chronic conditions.
2. Report the extent to which each of those levels is present among insured patients in a single market.
3. Compare patterns of care by type of insurance.
5
Data
• 4 years of claims data covering 1994-1997 in a single market.
• Patients covered by 5 different insurance plans:– 1 private indemnity plan– 2 private managed care plans– Medicare risk plan– Medicare indemnity plan
6
• Claims from the 5 plans were merged in a single dataset.
• Data on a final sample of 80,000 patients with 1 of 3 chronic conditions:– Asthma– Diabetes– Congestive Heart Failure (CHF).
7
Assumptions
1. For each condition, an optimal pattern of care for most patients exists.
2. Patients may obtain some, but not all of the recommended services.
Therefore, several different grades of quality can be described.
8
Example: Diabetes
5 components of good quality care:
1. Having an outpatient visit during the year.2. Receiving an HbA1c test during the year.3. Having continuity in hypoglycemic
medications for more than 6 months.4. Avoiding an ER visit or a diabetes-related
hospitalization.
9
5. Receive 3 or more of the following:a. Flu shot,
b. Podiatry visit,
c. Ophthalmology visit, or
d. Microalbumin test or on ARB or ACE Inhibitor.
10
5 Levels of Quality
1. Neglected care
2. Suboptimal care
3. Basic care
4. Better care
5. Best care
11
Outpatient Visit
Yes
No
Yes
Yes Yes
No
No
No
No
HbA1c Test HbA1c Test
ER Visit orHosp. Admit
PrescriptionContinuity
Better Basic NeglectedSuboptimalBest
FPOA*
PrescriptionContinuity
Yes Yes
Yes
No No
PrescriptionContinuity
FPOA*FPOA*
3+ <3<3 <33+ 3+
*F=Flu Shot; P=Podiatry Visit; O=Ophthalmology Visit; A = microalbumin test or ARB or ACE inhibitor
12
The Study Group
1. All patients with one of three chronic conditions in previous year.
2. Claims on at least two dates of service.
3. New to coverage in 1994.
4. Covered in all years, 1994 through 1997.
5. Data presented on calendar year basis.
13
Table 3 -- Quality categories for people newly eligible in 1994 and continuously eligible
from 1994 through 1997
1994 1995 1996 1997
a -- Asthma patients Neglected Care
36.5
34.4
33.6
33.9 Suboptimal Care 3.0 3.8 4.2 4.4 Basic Care 3.2 3.1 2.9 3.0 Better Care 52.8 49.5 47.4 45.5 Best Care 4.6 9.2 12.0 13.2
b -- CHF Patients Neglected Care
54.3
50.6
51.1
51.9 Suboptimal Care 7.6 8.6 9.5 8.3 Basic Care 29.6 26.2 25.6 24.0 Better or Best Care
8.5 14.6 13.8 15.8
c -- Diabetes patients Neglected Care
15.9
14.6
15.1
15.8 Suboptimal Care 58.8 56.7 52.8 50.2 Basic Care 21.1 22.5 23.9 23.7 Better or Best Care
4.1 6.3 8.1 10.3
N = Asthma: 5840; CHF: 2885; Diabetes: 11,505
14
Many patients were in the lowest categories:
• For diabetes and CHF, 59% to 71% were in lowest categories,
• For asthma patients, about 40% were in lowest categories.
15
Too few patients were in the highest categories:
• Only for asthma were more than half of patients in the top two categories.
• Fewer than 16% of patients in the other 2 categories were in the top two categories.
16
Table 4 – For patients continuously eligible from 1994 through 1997, the extent to which quality
category in 1994 predicts quality category in 1995
1994 1995
Neglected Care
Suboptimal Care
Basic Care
Better Care
Best Care
N
a. Asthma patients
Neglected Care 82.6 29.5 4.3 6.1 1.5 2010 Suboptimal Care 4.4 60.7 2.2 .5 1.9 222 Basic Care 1.0 .6 17.3 4.0 1.1 179 Better Care 10.5 3.5 67.6 79.4 33.5 2892 Best Care 1.6 5.8 8.6 10.1 62.1 537
N 2131 173 185 3082 269 5840 1994
1995
b. CHF patients
Neglected Care 85.4 16.0 9.4 3.1 … 1460 Suboptimal Care 5.0 43.8 6.9 6.3 4.3 249 Basic Care 7.3 25.1 60.0 32.3 8.7 756 Better Care 2.1 14.6 22.0 50.2 56.5 378 Best Care .1 .5 1.8 8.1 30.4 42
N 1566 219 854 223 23 2885 1994
1995
c. Diabetes patients Neglected Care 50.5 6.9 2.6 .5 1331 Suboptimal Care 42.0 79.3 22.1 17.5 6853 Basic Care 5.7 8.5 51.3 15.3 1899 Better or Best Care 1.7 5.3 24.0 66.6 1422
N 1549 7057 2129 770 11505
17
Trends
In most cases, having started out in a category, patients continued in that category from one year to the next.
18
Table 5 – For diabetes patients continuously eligible from 1994 through 1997, the extent to which
quality category in one year predicts quality category in the next year
1994 1995
Neglected Care
Suboptimal Care
Basic Care
Better or Best Care
N
Neglected Care 50.5 6.9 2.6 .5 1331 Suboptimal Care 42.0 79.3 22.1 17.5 6853 Basic Care 5.7 8.5 51.3 15.3 1899 Better or Best Care 1.7 5.3 24.0 66.6 1422
1996
Neglected Care 46.4 8.8 3.1 1.0 1417 Suboptimal Care 40.2 73.5 20.5 16.1 6372 Basic Care 9.4 9.1 41.1 13.2 1765 Better or Best Care 4.0 8.5 35.2 69.6 1951
1997
Neglected Care 43.4 10.0 3.2 1.4 1461 Suboptimal Care 41.1 69.9 19.2 13.8 6084 Basic Care 7.7 8.3 30.9 12.6 1457 Better or Best Care 7.8 11.8 46.7 72.2 2503 N 1549 7057 2129 770 11505
19
Table 6 -- Patients newly eligible in 1994 and continuously eligible from 1994 through 1997 by
quality category and insurance type
1994 1995 1996 1997 Indem MC Indem MC Indem MC Indem MC a. Asthma patients Neglected Care
62.1
20.7
59.0
19.3
57.8
18.6
57.8
19.2 Suboptimal Care 5.2 1.6 7.0 1.8 7.9 1.9 8.2 2.0 Basic Care 3.7 2.8 3.8 2.6 3.2 2.7 3.5 2.7 Better Care 27.7 68.3 24.0 65.3 24.6 61.5 23.1 59.3 Best Care 1.3 6.7 6.1 11.1 6.4 15.4 7.4 16.7
N 2230 3610 2228 3612 2215 3625 2215 3625 b. CHF patients Negligent
66.1
13.4
62.7
8.5
63.0
9.6
63.7
11.0 Suboptimal 8.7 3.6 10.4 2.5 11.1 4.0 9.1 5.6 Basic 22.1 55.6 18.0 54.8 18.0 52.2 18.1 44.6 Better or Best 3.1 27.5 8.9 34.2 7.9 34.2 9.1 38.8
N 2242 644 2240 646 2219 667 2211 675 c. Diabetes patients Negligent
18.4
7.2
17.3
4.3
18.8
4.3
19.4
4.5 Suboptimal 76.9 41.4 76.1 38.5 73.6 32.9 73.0 28.2 Basic 3.9 37.1 4.1 32.3 4.0 29.3 3.6 23.7 Better or Best .7 14.3 2.5 25.0 3.6 33.5 3.9 43.6
N 6456 5049 6448 5057 6356 5149 6337 5168
20
Insurance Type
In every year and for all three diagnoses, indemnity patients were much more likely to be in the lowest categories, and managed care patients were much more likely to be in the highest categories.
21
Limitations
• Administrative data – not a full picture of the clinical phenomenon of quality care.
• For indemnity patients, some utilization may be missing from the dataset.
• Claims may under-report services actually used.
• Data were from a single market. May not generalize to U.S.
22
Discussion
1. Large proportions of patients had utilization patterns that failed to meet quality criteria.
2. Patterns tended to persist from year to year.
23
Patients with Indemnity Insurance
3. Patients with indemnity insurance were much less likely to meet quality standards than patients in managed care.
4. Cost sharing is an obstacle to achieving optimal patterns of care.
5. Cost sharing reduces demand and, therefore, expenditures, but does not discriminate between needed and frivolous care.
24
Patients with Managed Care Coverage
6. For managed care patients, the numbers with appropriate patterns of care are too low, as well.
7. It is fair to say that MCOs and their physicians are not maximizing their opportunity to create optimal patterns of care.
25
The Challenge
Find methods to avoid unnecessary utilization while increasing, not reducing, the numbers with quality of care.
Use the advantages of managed care to increase the numbers of patients with appropriate patterns of care.