The project team used data collected during abstraction of patient records to compare the characteristics of SROS respondents and nonrespondents, as presented in Table 2-3.
|
Table 2-3. Comparisons of sample respondents/nonrespondents, using data from administrative records of cooperating providers (N=2,770)
|
|
Statistic
|
|
Respondents
|
|
Nonrespondents
|
|
|
(Base n max= 1,799)
|
|
(Base n max= 971)
|
|
Panel 1. Means
|
|
m
|
n
|
|
m
|
n
|
|
Length of sample episode (in months)
|
|
4.4
|
1,778
|
|
4.0
|
958
|
|
Age at admission (in years)
|
*
|
30.6
|
1,747
|
|
31.7
|
950
|
|
Number of different treatment services
received during this episode
|
|
3.3
|
1,698
|
|
3.5
|
920
|
|
Number of treatment episodes in lifetime
|
|
3.3
|
937
|
|
3.0
|
504
|
|
Panel 2. Percentages
|
|
%
|
n
|
|
%
|
n
|
|
Female
|
*
|
30.4
|
1,794
|
|
24.9
|
969
|
|
Black (non-Hispanic)
|
|
29.3
|
1,702
|
|
28.7
|
900
|
|
American Indian/Alaskan Native
|
|
1.5
|
1,702
|
|
1.0
|
900
|
|
White (non-Hispanic)
|
|
67.7
|
1,702
|
|
68.7
|
900
|
|
Hispanic
|
*
|
12.2
|
1,011
|
|
20.1
|
532
|
|
At least high school education
|
|
59.0
|
1,589
|
|
62.2
|
855
|
|
Prison or jail record prior to admission
|
|
49.7
|
1,026
|
|
49.7
|
563
|
|
DWI/DUI prior to admission
|
|
45.7
|
1,045
|
|
50.3
|
555
|
|
Other arrests prior to admission
|
|
70.7
|
1,222
|
|
68.2
|
682
|
|
Self as primary referral source
|
|
29.0
|
1,650
|
|
28.3
|
893
|
|
Legal system as primary referral
|
|
25.7
|
1,650
|
|
22.7
|
893
|
|
Public as primary payment source
|
|
28.8
|
1,439
|
|
32.0
|
762
|
|
Planned treatment greater than 25 days
|
|
69.0
|
598
|
|
67.2
|
344
|
|
Psychiatric history at admission
|
|
26.1
|
1,202
|
|
24.3
|
635
|
|
Employed at admission
|
|
39.1
|
1,621
|
|
37.9
|
890
|
|
Chronic medical condition at admission
|
|
29.2
|
1,254
|
|
33.2
|
660
|
|
Cocaine as primary drug at admission
|
|
15.5
|
1,360
|
|
17.0
|
746
|
|
Heroin as primary drug at admission
|
|
14.3
|
1,360
|
|
16.5
|
746
|
|
Alcohol as primary drug at admission
|
|
48.6
|
1,360
|
|
46.3
|
746
|
|
Prescribed medications during treatment
|
|
55.1
|
1,374
|
|
56.4
|
768
|
|
Length of treatment less than 25 days
|
|
64.0
|
1,778
|
|
60.6
|
954
|
|
Tested for drug or alcohol abuse during
sample episode
|
|
61.8
|
1,213
|
|
66.0
|
700
|
|
Ever used needles to inject drugs
|
*
|
43.5
|
928
|
|
49.9
|
521
|
|
Completed treatment plan before discharge
|
|
50.2
|
1,738
|
|
48.4
|
939
|
|
|
|
|
|
|
|
|
See footnotes at end of table
|
|
|
|
|
|
|
|
Percentages
|
|
%
|
n
|
|
%
|
n
|
|
Aftercare plan stated in record
|
|
65.6
|
1,132
|
|
67.8
|
599
|
|
Physician notes at admission
|
*
|
66.3
|
1,535
|
|
70.7
|
849
|
|
Physician notes at discharge
|
|
41.3
|
1,411
|
|
45.6
|
757
|
|
Physician notes at any other time
|
*
|
55.2
|
1,487
|
|
59.8
|
805
|
|
Hospital inpatient facility
|
|
24.5
|
1,799
|
|
21.6
|
971
|
|
Residential facility
|
|
25.8
|
1,799
|
|
28
|
971
|
|
Outpatient methadone maintenance facility
|
|
23.5
|
1,799
|
|
23.4
|
971
|
|
Outpatient non-methadone facility
|
|
26.2
|
1,799
|
|
27
|
971
|
|
Note: Significant differences between means in Panel 1 were based on two-tailed t tests.
Significant differences between percentages in Panel 2 were based on chi-square tests of
independence.
The table omits 277 cases from the total sample (3,047) who were known to be deceased
[i.e., 277 + 1,799 + 971 = 3,047].
* Difference between respondents and nonrespondents is significant, with p < .05.
Source: Drug Services Research Study/Supplemental Sample Abstraction conducted in 1994 at the 99 participating facilities.
|
The differences, though statistically significant, were small. An analysis of 36 variables collected by abstraction from client records — virtually all the data pertinent to client characteristics or individual treatment pathways — indicates very few significant differences (at the .05 level) between respondents and nonrespondents. Respondents averaged in age about one year younger (NS [ (NS) These differences are not significant at the .05 level when those who died are classified with respondents.] ) than nonrespondents, had a lower rate with physician notes at admission (NS), a lower rate with physician notes at any other time (NS), a higher rate of female respondents, a lower rate of Hispanic respondents, and a lower rate who used needles to inject drugs (NS). Given the small numbers and size of these differences, the study team did not adjust the data for nonresponse bias.
SAMHSA, an agency in the Department of Health and Human Services, is the Federal
Government's lead agency for improving the quality and availability of
substance abuse prevention, addiction treatment, and mental health
services in the United States.
