Cost effectiveness of interventions to improve adherence to statin therapy in ASCVD patients in the United States

Figures & data

Figure 1 Model schematics of decision tree and Markov transitions.

Table 1 Clinical inputs

Variable	Value	Confidence interval or range for sensitivity analysis	Distribution for sensitivity analysis	Source
Statin Treatment
Relative risk of coronary event year 1	0.86	0.77–0.95	Lognormal	4
Relative risk of revascularization year 1	0.95	0.84–1.08	Lognormal	4
Relative risk of stroke year 1	0.96	0.79–1.17	Lognormal	4
Relative risk of coronary event year 2	0.78	0.70–0.87	Lognormal	4
Relative risk of revascularization year 2	0.76	0.66–0.87	Lognormal	4
Relative risk of stroke year 2	0.75	0.62–0.90	Lognormal	4
Relative risk of coronary event year 2+	0.71	0.59–0.84	Lognormal	4
Relative risk of revascularization year 2+	0.73	0.59–0.9	Lognormal	4
Relative risk of stroke year 2+	0.79	0.57–1.1	Lognormal	4
Cardiovascular event distribution
CHD death	0.15	0.12–0.18	Dirichlet	27
Myocardial infarction	0.16	0.13–0.19	Dirichlet	27
Angina	0.38	0.30–0.46	Dirichlet	27
Revascularization	0.24	0.19–0.29	Dirichlet	27
Stroke	0.07	0.06–0.08	Dirichlet	27
Adherence
Probability of non-persistence year 1	0.23	0.19–0.28	Beta	6
Probability of non-persistence year 2	0.22	0.18–0.26	Beta	6
Probability of non-persistence year 3	0.16	0.13–0.19	Beta	6
Probability of non-compliance year 1	0.25	0.20–0.30	Beta	6,30
Probability of non-compliance year 2	0.25	0.20–0.30	Beta	6,30
Probability of non-compliance year 3	0.25	0.20–0.30	Beta	6,30
Probability of becoming compliant following MI	0.85	0.68–1.00	Beta	10
Probability of becoming compliant after CVD hospitalization	0.63	0.50–0.76	Beta	10
Probability of physician visit	0.70	0.56–0.84	Beta	10
Probability of becoming compliant following cardiologist visit	0.74	0.59–0.89	Beta	10
Probability of becoming compliant following primary care visit	0.58	0.46–0.69	Beta	10
Impact of adherence
Relative risk of CVD event with non-compliance vs compliance	1.35	1.21–1.50	Lognormal	15,19
Relative risk of all-cause mortality with non-compliance vs compliance	1.85	1.63–2.09	Lognormal	15,19
Relative risk of CVD event with non-persistence vs persistence	1.35	1.21–1.51	Lognormal	15,28
Relative risk of all-cause mortality with non-persistence vs persistence	2.78	1.96–3.72	Lognormal	15,28
Health state utilities
Starting utility secondary prevention	0.704	0.575–0.843	Beta	31
Post-second CVD event	0.581	0.452–0.720	Beta	31

Table 2 Cost inputs

Event (acute cost)	Value	Range	Distribution	Source
Non-fatal MI	$39,981	$27,615–$51,876	Gamma	23
Fatal MI	$52,267	$21,402–$59,943	Gamma	23
Cardiac arrest	$45,204	$17,137–$51,795	Gamma	23
Stroke	$74,057	$58,494–$94,048	Gamma	23
PCI	$84,821	$69,927–$113,597	Gamma	23
CABG	$162,911	$120,669–$268,096	Gamma	23
Angina	$19,978	$19,973–$19,982	Gamma	23
Cardiologist office visit	$146	$113–$160	Gamma	36
Primary care office visit	$74	$52–$81	Gamma	36
States (annual cost)
Compliant with statins	$132	$106–$158	Gamma	37
Non-compliant with statins	$79	$63–$95	Gamma	37
Secondary prevention state	$7,700	$5,828–$9,713	Gamma	32,38

Figure 2 Diagram of searches and study inclusion selection.

Table 3 Intervention summaries

Author	Intervention	Population	Comparator	Length of follow-up	Results
Coberly^Citation40	Disease Management (DM), unspecified	20,202 CHD patients in US administrative dataset	Before and after DM enrollment	12 months	43.9% with statin claim before and 45.6% after DM (p<0.0001)
Harbman^Citation43	Nurse Counseling, multiple time points using face-to-face and telephone; mean time per patient was 3.64 hrs	65 patients with recent MI at US hospital	Intervention vs usual care	3 months	86.7% adherence in usual care vs 100% adherence in intervention
Jelinek^Citation45	Nurse/Dietician Counseling, telephone-based risk management and medication coaching at multiple time points	656 patients discharged from coronary care or open heart surgical units at 2 Australian hospitals	Intervention vs usual care	24 months	87% adherence in usual care vs 94.4% with intervention (p<0.01)
Reddy^Citation47	Electronic Pill Bottle with Feedback	126 US Veterans with CAD	Intervention with direct feedback vs Intervention with family/friend feedback vs usual care	3 months	67% adherence in usual care vs 89% with direct feedback vs 86% with family/friend feedback (p<0.001)
Vollmer^Citation48	Interactive Voice Response (IVR) Reminders, IVR calls to overdue prescriptions or IVR calls + personalized health report to overdue prescriptions	25,323 US patients with ASCVD or diabetes mellitus and poor adherence	IVR vs IVR+ health report vs usual care	12 months	6.4% higher adherence in IVR group and 7.3% higher adherence in IVR+ group compared to usual care; Costs estimated at $9–$17 for IVR and $36–$47 for IVR+
Hohmann^Citation49	Discharge Letter, detailed discharge letter with rationale for all medications	312 ASCVD patients at a German hospital	Intervention vs usual care	3 months	69.8% of patients adherence in usual care vs 87.7% in intervention (p<0.001)

Table 4 Economic results

Intervention	Relative Improvement	Estimate Cost	Incremental Costs per patient	Incremental QALYs per patient	ICER
Disease Management^Citation40	1.04	$261.57 per patient annually	$2,424	0.02	$130,399
Nurse Counselling for 3.64 hrs every 3 mos.^Citation43	1.15	$745.37 per patient annually	$7,320	0.07	$111,173
Discharge Letter^Citation49	1.26	$51.19 once	$2,950	0.11	$27,545
Nurse/dietician Counselling^Citation45	1.11	$65.40 per patient annually	$1,808	0.05	$36,463
Electronic pill bottle + feedback^Citation47	1.33	$28.96 once for pill bottle + $40.86 per patient annually	$3,900	0.13	$29,631
IVR+^Citation48	1.16	$49.52 per patient annually	$2,169	0.07	$32,357

Figure 3 One-way sensitivity analysis.

Figure 4 Cost effectiveness plane for electronic pill bottle intervention.

Figure 5 Cost effectiveness acceptability curves for interventions.

Table S1 Search terms for structured literature review

Search categories (connected with Boolean AND)	List of Terms (connected with Boolean OR)	List of Subject Headings
Statin	Statin OR lipid lowering therapy OR lipid lowering treatment OR HMG-CoA reductase inhibitor	Hydroxymethylglutaryl-CoA Reductase Inhibitors/
Secondary Prevention	Cardiovascular disease OR CVD OR coronary artery disease OR coronary heart disease OR myocardial infarct* OR heart* OR cerebrovascular accident OR stroke* OR vascular disease	Cardiovascular diseases/ or exp heart diseases/ or exp vascular diseases/
Adherence	Adherence OR compliance OR persistence	Medication Adherence/
Cost	Cost* OR cost effectiveness OR cost utility analysis OR cost benefit analysis OR cea OR cua OR cba OR economic* OR economic evaluation	Costs and cost analysis/

Table S2 Inclusion/Exclusion criteria for literature review

Population	Patients with ASCVD
Intervention	Interventions included anything focused on improving patient adherence to statin therapy, including patient education, provider education, and prescription remainders. The review excluded interventions focused only changing patient costs for medication, new pill formulations and interventions focused on medications other than statins.
Comparator	Comparison groups included either a control arm of patients not receiving the intervention, or a historic control group of the patients receiving the intervention in the period before the intervention.
Outcomes	The primary outcome measured was change in medication adherence which may be defined as compliance, persistence or both, measured as proportion of days covered (PDC) or medication possession ratio (MPR), percent discontinuing treatment, time to discontinuation, or other methods. Secondary outcomes included impact of adherence on outcomes, either LDL-C level or second CVD event, as well as the cost of the intervention.

Figure S1 (A) Cost effectiveness plane IVR intervention. (B) Cost effectiveness plan of Disease Management intervention. (C) Cost effectiveness plane of Nurse Counseling intervention. (D) Cost effectiveness plane of Discharge Letter intervention. (E) Cost effectiveness plane of Nurse/Dietician Counseling intervention.

Coberly C, Morrow G, McGinnis M, et al. Increased adherence to cardiac standards of care during participation in cardiac disease management programs. Dis Manage. 2008;11(2):111–118. doi:10.1089/dis.2008.112725

 Google Scholar

Harbman P. The development and testing of a nurse practitioner secondary prevention intervention for patients after acute myocardial infarction: a prospective cohort study. Int J Nurs Stud. 2014;51:1542–1568. doi:10.1016/j.ijnurstu.2014.04.00424836930

 PubMed Web of Science ®Google Scholar

Jelinek M, Vale MJ, Liew D, et al. The COACH program produces sustained improvements in cardiovascular risk factors and adherence to recommended medications – two years follow-up. Heart Lung Circ. 2009;18:388–392. doi:10.1016/j.hlc.2009.06.00119648058

 PubMed Web of Science ®Google Scholar

Reddy A, Huseman TL, Canamucio A. Patient and partner feedback reports to improve statin medication adherence: a randomised control trial. J Gen Intern Med. 2016;32(3):256–261. doi:10.1007/s11606-016-3858-027612487

 PubMed Web of Science ®Google Scholar

Vollmer WM, Owen-Smith AA, Tom JO. Improving adherence to cardiovascular disease medications with information technology. Am J Manag Care. 2014;20(11):SP502–SP510.25811824

 PubMed Web of Science ®Google Scholar

Hohmann C, Neumann-Haefelin T, Klotz JM, Freidank A, Radziwill R. Providing systematic detailed information on medication upon hospital discharge as an important step towards improved transitional care. J Clin Ther. 2014;39:286–291.

 Web of Science ®Google Scholar