tradecraft CONTINUED
FIGURE 4
Quarterly Health Care Costs and Cost Differences—High-Risk Diagnosis Cohort
Quarter 1 Quarter 2 Quarter 3 Quarter 4 Quarter 5 Quarter 6 Quarter 7
Mean Quarterly Health Care Costs (IP, OP, RX)
Diagnosis Cohort: Cases of Depression $1,374.40
Diagnosis Cohort: Controls 1,263.32
P Value 0.179
Raw Difference
DiagnosisCohort—CasesMinusControls 111.08 9. 60 – 11. 27 80.15 337.99 134.72 553.16
Statistically Significant (5%) Results
Diagnosis Cohort—Cases Minus Controls Not Significant Not Significant Not Significant Not Significant 337.99 Not Significant 553.16
Quarter 8
$1,159.02
1,149.42
0.463
$1,218.28
1,229.55
0.461
$1,374.06
1,293.91
0.250
$1,588.03
1,250.04
0.003
$1,413.57
1,278.85
0.121
$1,712.18
1,159.92
<0.001
$1,852.53
1,142.46
<0.001
710.07
710.07
FIGURE 5
Quarterly Health Care Costs and Cost Differences—High-Risk Rx Cohort
Quarter 1 Quarter 2 Quarter 3 Quarter 4 Quarter 5 Quarter 6
Mean Quarterly Health Care Costs (IP, OP, RX)
Rx Cohort: Cases of Antidepressant Usage $1,250.37
Rx Cohort: Controls 1,189.99
P Value 0.209
Raw Difference
Rx Cohort—Cases Minus Controls 60. 38 65.02 36. 25 121.07 245.23 379.76
Statistically Significant (5%) Results
Rx Cohort—Cases Minus Controls Not Significant Not Significant Not Significant Not Significant 245.23 379.76
Quarter 7 Quarter 8
$1,340.43
1,275.41
0.212
$1,297.35
1,261.10
0.322
$1,386.48
1,265.41
0.064
$1,524.16
1,278.92
0.002
$1,672.63
1,292.87
<0.001
$1,685.15
1,314.00
<0.001
$2,602.11
1,174.32
<0.001
371.14
1,427.79
371.14
1,427.79
identified in the claims data. A person
was identified as depressed either by
presence of an ICD- 9 diagnostic code
(the diagnosis cohort) or by the prescription of an antidepressant (the Rx
cohort).
■ ■ Medical Risk Severity—We assigned
a risk score to each individual using
his or her cumulative medical risk
based on ICD- 9 codes present in the
medical claims data. Members were
stratified as high risk (worse than
average) and low risk (average or
better than average). This separation
allowed us to understand whether
there was a greater opportunity for
cost reduction in one group compared
with the other.
FIGURE 6
Significant Differences Between Cases and Controls:
High-Risk Group, Health Care Costs
$1,600
1,400
1,200
Diagnosis Cohort
Rx Cohort
1,000
While the excess health care and
absenteeism costs for people with undiagnosed depression over the two-year
period preceding diagnosis were approximately $3,386 per undiagnosed
depressed member, certain cohorts had
the largest cost differences in absence
costs. In the low-risk diagnosis cohort,
for example, approximately 74 percent
of the total excess costs were due to
absence from work. In the low-risk Rx
cohort, the absence-from-work costs
were approximately 58 percent of the
total. Other cohorts had the largest excess cost differences in health care costs.
In both the high-risk diagnosis and high-risk Rx cohorts, excess health care costs
were the dominant contributor to total
excess costs, constituting between 62
percent and 67 percent of the total.
800
600
400
200
0
–200
Quarter 1
Quarter 2
Quarter 3
Period Prior to Diagnosis
Quarter 4 Quarter 5
Quarter 6
Quarter 7
Quarter 8
Health Care Costs and
Depression
Figures 4, 5, 7, and 8 compare the quar-
terly total health care cost (inpatient,
outpatient, and pharmacy) between
cases and controls, with Figures 4 and
5 detailing the results for high-risk
groups in the diagnosis and Rx cohorts
and Figures 7 and 8 showing the results
for the low-risk groups in the diagnosis
66 CONTINGENCIES JUL | AUG. 12
WWW.CONTINGENCIES.ORG
