CMS-10709. 340B Survey Characteristics Non-Response Bias Report

340B Characteristics Nonresponse Bias Report
(CMS-10709; 0938-1374)
HOSPITAL CHARACTERISTICS ANALYIS
We are providing this nonresponse bias report to the Office of Information and Regulatory Affairs (OIRA)
prior to our utilization of the 340B drug acquisition cost data for future publications, including rulemaking,
consistent with OIRA’s terms of clearance of approval for the 340B survey. While OIRA used the term
“nonresponse bias” we note that the survey was designed such that lack of an affirmative response to the
survey in whole or in part would lead CMS to use the 340B ceiling price as a proxy for the acquisition
costs of applicable drugs.
The content of this report describes the scope and characteristics of the responding hospitals such as
bed count, teaching hospital status, hospital type, and geographic locality type, as well as any limitations
in the generalizability of the information collected, in any publications and documents utilizing this data.
This report summarizes the survey responses by stratifying by several hospital and geographic
characteristics including: bed count, teaching hospital status, hospital type, and geographic locality type.
We analyzed which hospitals completed the detailed survey, opted for the quick survey, or did not
respond at all.
Following the OPPS rate changes in January 2018 for drugs acquired under the 340B Drug Pricing
Program and litigation brought by certain hospitals, CMS began re-evaluating the average hospital
acquisition costs for 340B-acquired specified outpatient drugs and whether the agency could gather such
information via a survey. To do this, a survey on acquisition drug costs was administered to 340B covered
entity hospitals with the aim of using the data to ascertain these costs. CMS administered the survey to
1,422 340B entities between April 24th and May 15th, 2020. There were two survey options available to
responders:



Detailed Survey: Hospitals that completed the detailed survey were required to provide their
average acquisition costs for each individual specified covered outpatient drug.
Quick Survey: Hospitals that completed the quick survey opted for CMS to use the HRSA 340B
ceiling prices to be reflective of their own acquisition costs.

DESCRIPTIVE STATISTICS SUMMARY
Out of the 1,422 340B covered entity hospitals that were surveyed, 100 (7.0%) completed the detailed
survey, 780 (54.9%) completed the quick survey, and 542 (38.1%) did not respond. Thirty-three percent
of the detailed survey responders, 25.6 percent of the quick survey responders, and 34.3 percent of the
non-responders had a bed count of less than 99 beds.
Table 1.

Count
Percent

340B
Hospitals
Surveyed
1422
100.0%

Detailed
Survey
Responses
100
7.0%

Quick
Survey
Responses
780
54.9%

No
Response
542
38.1%

In the aggregate population of 340B hospitals, 51.3 percent are teaching hospitals and 45.5 percent are
non-teaching. Fifty-two percent of the detailed survey responders, 55.5 percent of the quick survey

1

responders, and 45.0 percent of the non-responders were teaching hospitals. A full listing of this
information is included in Appendix A.
The majority (79.0%) of 340B covered entity hospitals that were surveyed are disproportionate share
(DSH) hospitals. Of the detailed responders, 67.0 percent were DSH hospitals, 23.0 percent were sole
community hospitals (SCH), 8.0 percent were rural referral center (RRC) hospitals, and 1.0 percent were
children’s hospitals. Of the quick responders, 80.3 percent were DSH, 9.5 percent were SCH, and 5
percent were RRC. Seventy-nine point five percent of the entities that did not respond to the survey were
DSH, and 10.7 percent were SCH.
The aggregate population of 340B covered entity hospitals, all of which were surveyed, is located
primarily in urban areas and has concentrations in the East North Central, South Atlantic, and Pacific
regions1. Fifty-nine percent of the detailed responders were hospitals from urban areas, and 35 percent
were from the East North Central region. Seventy-five percent of the quick responders were from urban
areas, and 20.4 percent from the East North Central region. The majority of detailed responses were from
Michigan, Ohio, and California, while the majority of quick responses were from New York, California, and
Illinois. With 89 entities that did not complete the survey, California also had the highest non-response
rate.
STATISTICAL ANALYSIS
Using a chi-squared test, the aggregate detailed and quick survey responders is not statistically
significantly different from the overall population of 340B hospitals for the majority of characteristics with a
95% confidence level (i.e., bed count, teaching hospital status, entity type, and urban/rural location).
However, the survey responders differ in terms of geographic region from the overall population.
Additionally, the population of survey responders is statistically significantly different from the overall
population of 340B hospitals in terms of 340B drug claim line volume at a 95% confidence level but not at
a 99% confidence level (p-value = 0.0121). Please see Appendix B, Table 4.
The chi-squared summaries for the detailed and quick survey responders are available in Appendix B,
Tables 5 and 6, respectively.
CONCLUSION
In conclusion, about 61 percent of the 340B hospitals submitted a quick or detailed survey. The
characteristics of hospitals that completed the quick or detailed survey were similar to those observed in
the overall population of 340B hospitals in terms of entity type, bed size, 340B drug volume, geographic
locality, and teaching hospital status. For example, survey responses were received from cancer
hospitals, children’s hospitals, sole community hospitals and RRCs in addition to a large number of DSH
hospitals. Therefore, we do not believe there are limitations in the generalizability of the survey data to
the population of 340B covered entities.

1

For a detailed list of abbreviations and states assigned to regions see Appendix A, Table 3.

2

Appendix A.
Table 1. Hospital Characteristics - Summary of 340B Survey Responders
# 340B
Entities
Bed Count
Total

% 340B
Entities

# Detailed
Survey

% Detailed
Survey

# Quick
Survey

% Quick
Survey

# No
Affirmative
Response

% No
Affirmative
Response

1,422

100.0%

100

100.0%

780

100.0%

542

100.0%

419
347

29.5%
24.4%

33
18

33.0%
18.0%

200
178

25.6%
22.8%

186
151

34.3%
27.9%

200 - 299
214
300 - 499
232
500 +
164
Missing
46
Teaching Hospital
Total
1,422
Teaching
729
Non-Teaching
647

15.0%
16.3%
11.5%
3.2%

21
20
8
0

21.0%
20.0%
8.0%
0.0%

123
132
115
32

15.8%
16.9%
14.7%
4.1%

70
80
41
14

12.9%
14.8%
7.6%
2.6%

100.0%

100

100.0%

780

100.0%

542

100.0%

51.3%

52

52.0%

433

55.5%

244

45.0%

45.5%

48

48.0%

315

40.4%

284

52.4%

3.2%

0

0.0%

32

4.1%

14

2.6%

100

100.0%

780

100.0%

542

100.0%

0 - 99
100 - 199

Missing

46

340B Drug Claim Line Volume (CY 2018)
Total
1,422
100.0%
0 - 499
522
36.7%

32

32.0%

241

30.9%

249

45.9%

115
458
152
158
17

8.1%
32.2%
10.7%
11.1%
1.2%

10
34
14
10
0

10.0%
34.0%
14.0%
10.0%
0.0%

57
267
95
106
14

7.3%
34.2%
12.2%
13.6%
1.8%

48
157
43
42
3

8.9%
29.0%
7.9%
7.7%
0.6%

Hospital Type
Total

1,422

100.0%

100

100.0%

780

100.0%

542

100.0%

CAN
DSH

3
1,124

0.2%
79.0%

0
67

0.0%
67.0%

2
626

0.3%
80.3%

1
431

0.2%
79.5%

500 - 999
1,000 - 4,999
5,000 - 9,999
10,000 - 49,999
50,000 +

3

PED
RRC
SCH
Missing

51
74
155

Rural
Urban
Region

395
1027

Total
East North
Central

3.6%
5.2%
10.9%

1
8
23

1.0%
8.0%
23.0%

37
39
74

4.7%
5.0%
9.5%

13
27
58

2.4%
5.0%
10.7%

1.1%

1

1.0%

2

0.3%

12

2.2%

100.0%

100

100.0%

780

100.0%

542

100.0%

27.8%
72.2%

41
59

41.0%
59.0%

196
584

25.1%
74.9%

158
384

29.2%
70.8%

1,422

100.0%

100

100.0%

780

100.0%

542

100.0%

228

16.0%

35

35.0%

159

20.4%

34

6.3%

133

9.4%

5

5.0%

69

8.8%

59

10.9%

171

12.0%

11

11.0%

113

14.5%

47

8.7%

89
61
223
5

6.3%
4.3%
15.7%
0.4%

3
5
10
0

3.0%
5.0%
10.0%
0.0%

51
39
78
0

6.5%
5.0%
10.0%
0.0%

35
17
135
5

6.5%
3.1%
24.9%
0.9%

227

16.0%

12

12.0%

123

15.8%

92

17.0%

West North
Central

112

7.9%

7

7.0%

67

8.6%

38

7.0%

West South
Central

173

12.2%

12

12.0%

81

10.4%

80

14.8%

15
Geographic Locality Type
Total
1,422

East South
Central
Middle
Atlantic
Mountain
New England
Pacific
Puerto Rico
South Atlantic

Table 2: State Summary of Survey Responses

4

State
Total

# 340B
Entities

% 340B
Entities

# Full
Survey

% Full
Survey

# Quick
Survey

% Quick
Survey

# No
Response

% No
Response

1,422

100%

100

100%

780

100%

542

100%

California

150

10.5%

7

7.0%

54

6.9%

89

16.4%

New York

95

6.7%

5

5.0%

72

9.2%

18

3.3%

Texas

83

5.8%

6

6.0%

40

5.1%

37

6.8%

Illinois

63

4.4%

5

5.0%

46

5.9%

12

2.2%

Ohio

61

4.3%

11

11.0%

38

4.9%

12

2.2%

Pennsylvania

54

3.8%

5

5.0%

30

3.8%

19

3.5%

Georgia

53

3.7%

5

5.0%

26

3.3%

22

4.1%

Michigan

51

3.6%

11

11.0%

35

4.5%

5

0.9%

North Carolina

51

3.6%

0

0.0%

33

4.2%

18

3.3%

Louisiana

47

3.3%

3

3.0%

18

2.3%

26

4.8%

Kentucky

44

3.1%

1

1.0%

30

3.8%

13

2.4%

Florida

39

2.7%

1

1.0%

30

3.8%

8

1.5%

Missouri

35

2.5%

3

3.0%

23

2.9%

9

1.7%

Washington

35

2.5%

1

1.0%

13

1.7%

21

3.9%

Indiana

32

2.3%

6

6.0%

24

3.1%

2

0.4%

Alabama

32

2.3%

1

1.0%

13

1.7%

18

3.3%

Mississippi

31

2.2%

2

2.0%

10

1.3%

19

3.5%

Colorado

29

2.0%

0

0.0%

21

2.7%

8

1.5%

Minnesota

27

1.9%

0

0.0%

20

2.6%

7

1.3%

Maryland

26

1.8%

2

2.0%

4

0.5%

20

3.7%

Tennessee

26

1.8%

1

1.0%

16

2.1%

9

1.7%

Massachusetts

26

1.8%

1

1.0%

18

2.3%

7

1.3%

Oregon

25

1.8%

2

2.0%

4

0.5%

19

3.5%

Arizona

25

1.8%

1

1.0%

11

1.4%

13

2.4%

Iowa

22

1.5%

4

4.0%

8

1.0%

10

1.8%

5

Arkansas

22

1.5%

3

3.0%

10

1.3%

9

1.7%

New Jersey

22

1.5%

1

1.0%

11

1.4%

10

1.8%

Wisconsin

21

1.5%

2

2.0%

16

2.1%

3

0.6%

Oklahoma

21

1.5%

0

0.0%

13

1.7%

8

1.5%

South Carolina

17

1.2%

1

1.0%

10

1.3%

6

1.1%

Virginia

17

1.2%

0

0.0%

14

1.8%

3

0.6%

West Virginia

14

1.0%

2

2.0%

2

0.3%

10

1.8%

Connecticut

12

0.8%

2

2.0%

10

1.3%

0

0.0%

Kansas

12

0.8%

0

0.0%

5

0.6%

7

1.3%

New Mexico

11

0.8%

0

0.0%

7

0.9%

4

0.7%

Maine

10

0.7%

2

2.0%

7

0.9%

1

0.2%

Hawaii

9

0.6%

0

0.0%

6

0.8%

3

0.6%

Montana

8

0.6%

0

0.0%

2

0.3%

6

1.1%

Utah

7

0.5%

0

0.0%

5

0.6%

2

0.4%

Idaho

6

0.4%

2

2.0%

4

0.5%

0

0.0%

Delaware

6

0.4%

1

1.0%

2

0.3%

3

0.6%

Nebraska

6

0.4%

0

0.0%

2

0.3%

4

0.7%

Vermont

6

0.4%

0

0.0%

3

0.4%

3

0.6%

South Dakota

6

0.4%

0

0.0%

6

0.8%

0

0.0%

Puerto Rico

5

0.4%

0

0.0%

0

0.0%

5

0.9%

Rhode Island

4

0.3%

0

0.0%

0

0.0%

4

0.7%

Alaska

4

0.3%

0

0.0%

1

0.1%

3

0.6%

District of Columbia

4

0.3%

0

0.0%

2

0.3%

2

0.4%

North Dakota

4

0.3%

0

0.0%

3

0.4%

1

0.2%

New Hampshire

3

0.2%

0

0.0%

1

0.1%

2

0.4%

Nevada

2

0.1%

0

0.0%

1

0.1%

1

0.2%

Wyoming

1

0.1%

0

0.0%

0

0.0%

1

0.2%

6

Table 3: Key
Region Abbreviations
ESC = East South Central
MA = Middle Atlantic
SA = South Atlantic
WSC = West South Central
WNC = West North Central
MNT = Mountain
ENC = East North Central
PAC = Pacific
NE = New England
PR = Puerto Rico
STATE
ALABAMA
ALASKA
ARIZONA
ARKANSAS
CALIFORNIA
COLORADO
CONNECTICUT
DELAWARE
DISTRICT OF COLUMBIA
FLORIDA
GEORGIA
HAWAII
IDAHO
ILLINOIS
INDIANA
IOWA
KANSAS
KENTUCKY
LOUISIANA
MAINE

REGION
ESC
PAC
MNT
WSC
PAC
MNT
NE
SA
SA
SA
SA
PAC
MNT
ENC
ENC
WNC
WNC
ESC
WSC
NE

MARYLAND
MASSACHUSETTS
MICHIGAN
MINNESOTA
MISSISSIPPI
MISSOURI
MONTANA
NEBRASKA
NEVADA
NEW HAMPSHIRE
NEW JERSEY
NEW MEXICO
NEW YORK
NORTH CAROLINA
NORTH DAKOTA
OHIO
OKLAHOMA
OREGON
PENNSYLVANIA
PUERTO RICO
RHODE ISLAND

SA
NE
ENC
WNC
ESC
WNC
MNT
WNC
MNT
NE
MA
MNT
MA
SA
WNC
ENC
WSC
PAC
MA
PR
NE

SOUTH CAROLINA
SOUTH DAKOTA
TENNESSEE
TEXAS
UTAH
VERMONT
VIRGIN ISLANDS
VIRGINIA
WASHINGTON
WEST VIRGINIA
WISCONSIN

SA
WNC
ESC
WSC
MNT
NE
#N/A
SA
PAC
SA
ENC

7

WYOMING
MNT
CALIFORNIA
PAC
AMERICAN SAMOA
GUAM
NORTHERN MARIANAS ISLANDS
TEXAS
WSC
FLORIDA
SA
FLORIDA
SA
KANSAS
WNC
LOUISIANA
WSC
OHIO
ENC
PENNSYLVANIA
MA
TEXAS
WSC
MARYLAND
SA

Appendix B.
Table 4. Chi-Square Summary of Aggregate 340B Survey Responders
%
Responders

% 340B
Hospitals

Chi-Square
p-value

Chi-Square
Interpretation

Bed Count
0 - 99

26.5%

29.5%

100 - 199

22.3%

24.4%

200 - 299

16.3%

15.0%

300 - 499

17.3%

16.3%

500 +

14.0%

11.5%

3.6%

3.2%

Teaching

55.1%

51.3%

Non-Teaching

41.3%

45.5%

3.6%

3.2%

Missing

0.2584

The population of responders is not statistically significantly
different from the overall population of 340B hospitals.

0.1339

The population of responders is not statistically significantly
different from the overall population of 340B hospitals.

0.0121

The population of responders is statistically significantly
different from the overall population of 340B hospitals at the
0.05 level. However, the population of responders is not
statistically significantly different from the overall population
of 340B hospitals at the 0.01 level.

0.5022

The population of responders is not statistically significantly
different from the overall population of 340B hospitals.

Teaching Hospital

Missing

340B Drug Claim Line Volume (CY 2018)
0 - 499

35.0%

41.6%

8.5%

9.3%

1,000 - 4,999

35.9%

32.3%

5,000 - 9,999

11.4%

9.8%

9.2%

7.1%

CAN

0.2%

0.2%

DSH

78.8%

79.0%

PED

4.3%

3.6%

RRC

5.3%

5.2%

SCH

11.0%

10.9%

0.3%

1.1%

500 - 999

10,000 - 49,999
Hospital Type

Missing

8

Geographic Locality Type
Rural

26.9%

27.8%

Urban

73.1%

72.2%

22.1%

16.0%

8.4%

9.4%

14.1%

12.0%

Mountain

6.1%

6.3%

New England

5.0%

4.3%

10.0%

15.7%

Puerto Rico

0.0%

0.4%

South Atlantic
West North
Central
West South
Central

15.3%

16.0%

8.4%

7.9%

10.6%

12.2%

Region
East North
Central
East South
Central
Middle Atlantic

Pacific

0.6585

The population of responders is not statistically significantly
different from the overall population of 340B hospitals.

0.0003

The population of responders is statistically different from
the overall population of 340B hospitals.

9

Table 5. Chi-Square Summary of 340B Detailed Survey Responders
% Full
Responders

% 340B
Entities

Chi-Square
p-value

Chi-Square
Interpretation

Bed Count
0 - 99

33.0%

29.5%

100 - 199

18.0%

24.4%

200 - 299

21.0%

15.0%

300 - 499

20.0%

16.3%

500 +

8.0%

11.5%

Missing

0.0%

3.2%

Teaching

52.0%

51.3%

Non-Teaching

48.0%

45.5%

0.0%

3.2%

0.1026

The population of responders is not statistically significantly
different from the overall population of 340B hospitals.

0.1852

The population of responders is not statistically significantly
different from the overall population of 340B hospitals.

0.9538

The population of responders is not statistically significantly
different from the overall population of 340B hospitals.

0.0048

The population of responders is statistically different from
the overall population of 340B hospitals.

Teaching Hospital

Missing

340B Drug Claim Line Volume (CY 2018)
0 - 499

40.0%

41.6%

9.0%

9.3%

1,000 - 4,999

36.0%

32.3%

5,000 - 9,999

9.0%

9.8%

10,000 - 49,999

6.0%

7.1%

CAN

0.0%

0.2%

DSH

67.0%

79.0%

PED

1.0%

3.6%

RRC

8.0%

5.2%

SCH

23.0%

10.9%

1.0%

1.1%

500 - 999

Hospital Type

Missing

Please note that the Chi-Square analysis may not be the
most appropriate statistical test for this measure due to low
sample size since 33% of the categories have expected
counts less than 5.

Geographic Locality Type

10

Rural

41.0%

27.8%

Urban

59.0%

72.2%

35.0%

16.0%

5.0%

9.4%

11.0%

12.0%

3.0%

6.3%

Region
East North
Central
East South
Central
Middle Atlantic
Mountain
New England

5.0%

4.3%

10.0%

15.7%

Puerto Rico

0.0%

0.4%

South Atlantic
West North
Central
West South
Central

12.0%

16.0%

7.0%

7.9%

12.0%

12.2%

Pacific

0.0047

0.0016

11

The population of responders is statistically different from
the overall population of 340B hospitals.
The population of responders is statistically different from
the overall population of 340B hospitals.

Table 6. Chi-Square Summary of 340B Quick Survey Responders
% Quick
Responders

% 340B
Entities

Chi-Square
p-value

Chi-Square
Interpretation

Bed Count
0 - 99

25.6%

29.5%

100 - 199

22.8%

24.4%

200 - 299

15.8%

15.0%

300 - 499

16.9%

16.3%

500 +

14.7%

11.5%

4.1%

3.2%

Teaching

55.5%

51.3%

Non-Teaching

40.4%

45.5%

4.1%

3.2%

Missing

0.1261

The population of responders is not statistically significantly
different from the overall population of 340B hospitals.

0.0548

The population of responders is not statistically significantly
different from the overall population of 340B hospitals.

0.0053

The population of responders is statistically different from
the overall population of 340B hospitals.

0.2211

The population of responders is not statistically significantly
different from the overall population of 340B hospitals.

Teaching Hospital

Missing

340B Drug Claim Line Volume (CY 2018)
0 - 499

34.4%

41.6%

8.5%

9.3%

1,000 - 4,999

35.9%

32.3%

5,000 - 9,999

11.7%

9.8%

9.6%

7.1%

CAN

26.0%

0.2%

DSH

80.3%

79.0%

PED

4.7%

3.6%

RRC

5.0%

5.2%

SCH

9.5%

10.9%

Missing

0.3%

1.1%

500 - 999

10,000 - 49,999
Hospital Type

Geographic Locality Type

12

Rural

25.1%

27.8%

Urban

74.9%

72.2%

20.4%

16.0%

8.9%

9.4%

14.5%

12.0%

6.5%

6.3%

Region
East North
Central
East South
Central
Middle Atlantic
Mountain
New England

5.0%

4.3%

10.0%

15.7%

Puerto Rico

0.0%

0.4%

South Atlantic
West North
Central
West South
Central

15.8%

16.0%

8.6%

7.9%

10.4%

12.2%

Pacific

0.1796

The population of responders is not statistically significantly
different from the overall population of 340B hospitals.

0.0031

The population of responders is statistically different from
the overall population of 340B hospitals.

13