Tria Health

Question/ Intervention Goal/ Objectives

Tria Health’s goal is to reduce drug costs for chronic conditions that require multiple medications and to increase medication compliance. Better medication compliance has been shown to reduce healthcare costs, such as from hospital stays.

Method/ Calculation

The analysis reviewed emergency room (ambulatory) and inpatient hospital visits for selected chronic conditions. A group of 1,708 members who used the program were matched to 1,708 members who had not used the program. Matching was done based upon member age, diagnosis history, and quartile of per member per month medical costs. Diagnosis history focused on whether the member had claims with any of the following as a primary diagnosis: congestive heart failure, hypertension, asthma, chronic obstructive pulmonary disease, diabetes, and coronary artery disease. All members included in the analysis had continuous coverage in the benefit plan during the time periods reviewed.

The control (non-user) group’s year-over-year trend was applied to the intervention (user) group. The intervention group’s actual average cost per ER and inpatient visit were used to calculate cost savings.

Findings/ Metric/ Outcome/ Savings

See Table I below for a summary of the study group’s actual trends and comparison to the control group’s trends. The gross difference between the study group’s actual costs and the hypothetical costs was $782,104

Table 1 — Study Group Actuals/ Control Group Trends

2017 2018 2019
Emergency Room Visits
Actual ER Rate / 100 Members 1.46 1.46 1.93
$90,168 $102,471
Hypothetical ER Rate/ 100 members ^ 1.90 2.01
ER visits – hypothetical 32 34
Hypothetical Costs $116,885 $106,381
Actual vs. hypothetical $(26,716) $(3,910)
Inpatient Stays
Actual IP Rate/ 100 Members 1.81 1.35 1.11
Actual IP admits 31 23 19
Actual Costs $499,438 $379,091
Hypothetical IP rate/ 100 members ^ 2.88 1.65
IP visits hypothetical 49 28
Hypothetical Costs $1,069,130 $560,878
Actual – Hypothetical $(569,692) $(181,786)


The frequency of the target ER and inpatient visits is relatively low for a working-age population. The infrequency likely contributed to the intervention and control groups having different baseline rates and different average per-visit costs. The analysis strives for a conservative estimate of savings, by applying the (lower) study group’s average visit costs to the calculation. This may understate the true impact.