Cost and Value in Contemporary Heart Failure Clinical Guidance Documents

John W. Ostrominski, MD; Sameer Hirji, MD, MPH; Ankeet S. Bhatt, MD, MBA; Javed Butler, MD, MPH, MBA; Mona Fiuzat, PHARMD; Gregg C. Fonarow, MD; Paul A. Heidenreich, MD, MS; James L. Januzzi, JR, MD; Carolyn S.P. Lam, MBBS, PHD; Thomas M. Maddox, MD, MS; Christopher M. O'Connor, MD; Muthiah Vaduganathan, MD, MPH


JACC Heart Fail. 2022;10(1):1-11. 

In This Article


Document Characteristics

Overall, 33 HF CGDs were identified, including 5 (15%) appropriate use criteria, 7 (21%) clinical practice guidelines, and 21 (64%) expert consensus documents. All CGDs included in the analysis are detailed in Supplemental Table 1. Of 15 (45%) CGDs that included methodology sections, cost/value was mentioned in 7. In only 1 of these 7 instances was cost/value stated to be explicitly included in the development of clinical guidance recommendations. Cost/value was reported to be implicitly included in the 6 remaining methodology sections.

Eight (24%) documents included sections devoted to discussion of cost/value issues, of which 3 reviewed cost/effectiveness or cost/benefit of HF-related interventions, 3 appraised the economic impact of HF or HF-related care, and 2 discussed guidance for the management of patient-level costs of care. Three (9%) documents reported estimated costs of HF-related devices, only 1 of which estimated out-of-pocket cost. No documents reported estimated out-of-pocket costs for recommended HF-related pharmacological therapies.

Cost and Value Statements and Citations

A total of 179 statements related to cost/value were identified in 27 (82%) CGDs (median 3 [IQR: 1–6]) per document (Figure 1A). Most of these statements were found in expert consensus documents (77.7% vs 17.3% in clinical practice guidelines and 5.0% in appropriate use criteria). These patterns were preserved following accounting for the number of documents of each type (median 4 cost/value statements per expert consensus document vs median 3 per clinical practice guideline vs median 2 per appropriate use criteria). Twenty (61%) CGDs included at least 1 cost/value-related citation (overall median 1 [IQR: 0–3]) (Figure 1B), of which 34 (39%) represented cost-effectiveness or cost-utility analyses and 53 (61%) were descriptive.

Figure 1.

Cost/Value Statements and Citations per Document
(A) Most contemporary heart failure–focused clinical guidance documents (27 of 33; 82%) included at least 1 cost/value statement, with a median of 3 cost/value statements per document (IQR: 1–6). Most of these statements were found in expert consensus documents. (B) Slightly fewer clinical guidance documents (20 of 33; 61%) included at least 1 cost/value-related citation (overall median 1 [IQR: 0–3]).

More recently published CGDs had a frequency of cost/value statements similar to those published before 2016 (median 4 [IQR: 2–6] for documents published between 2010 and 2015 vs median 3 [IQR: 0.5–6.0] for documents published between 2016 and 2021; P = 0.56). Similar patterns were observed for inclusion of cost/value citations (median 2 [IQR: 0–3] for documents published between 2010 and 2015 vs median 0.5 [IQR: 0.0–2.5] for documents published between 2016 and 2021; P = 0.33). Figure 2 highlights the distribution of cost/value statements over time.

Figure 2.

All Cost/Value Statements by Statement Category, 2010 to 2021
Appreciable variation was observed in the number of cost/value statements over time, but more recently published clinical guidance documents had a similar frequency of cost/value statements as compared with those published before 2016 (P = 0.56). Cost/value statements highlighting the economic impact of heart failure were the most common overall, and they were the only cost/value statement type encountered during every year of the study interval.

Of the 179 statements, 116 (64.8%) highlighted the economic impact of HF or HF-related care, 6 (3.4%) advocated for cost/value issues, 15 (8.4%) reported gaps in cost/value evidence, and 42 (23.5%) supported clinical guidance recommendations. Examples of statement categories are presented in Supplemental Table 2. Economic impact statements remained most common across each document type (expert consensus documents, clinical practice guidelines, appropriate use criteria) (Figure 3). Expert consensus documents were the only document type to include cost/value statements from all 4 types of categories assessed.

Figure 3.

Distribution of Cost/Value Statements by Category, by Document Type
Expert consensus documents were the only document type to include cost/value statements from all categories.

Of the cost/value statements supporting clinical guidance recommendations, 19 (45.2%) focused on the management of societal cost, whereas 23 (54.8%) focused on patient-level cost. Of the statements focusing on patient-level costs, 18 were published between 2016 and 2021 as compared with 5 between 2010 and 2015, denoting an increasing focus on the management of patient-level cost over the study interval.

Of the 42 cost/value statements supporting clinical guidance recommendations, 38 (90.5%) supported use of an intervention, whereas 4 (9.5%) discouraged use of an intervention. Of those statements supporting intervention use, 33 recommended use to avoid future costs (eg, HF education before discharge and continuous home ambulatory inotropic agent infusions for patients with end-stage HF), 3 recommended use citing equal effectiveness at lower cost in routine cases (eg, transthoracic echocardiography [TTE] compared with cardiac magnetic resonance [CMR] as first-line cardiovascular imaging modality), and 2 recommended use because the incremental benefit justified additional cost (eg, cardiac resynchronization therapy in selected patients with HF and aerobic training programs for patients with HF in skilled nursing facilities). A total of 29 of the 38 statements recommending use were not accompanied by any specific supporting evidence. Of the remaining 9 statements, 2 were supported by cost-effectiveness analyses of randomized trials (for cardiac resynchronization therapy and comprehensive HF disease management programs), and 1 was supported by a preliminary cost analysis of a small pilot trial (observation unit-based HF management). All others were supported by observational evidence.

All 4 statements discouraging use referred to diagnostic interventions; there were no examples of a statement supporting nonuse of a specific therapeutic intervention because of cost/value. Of these 4 statements, 3 supported nonuse of screening initiatives for asymptomatic left ventricular systolic dysfunction; routine use of TTE and/or biomarker-based screening was discouraged, citing uncertain cost/benefit as a result of insufficient prospective cost-effectiveness data. The 1 remaining "nonuse" statement supported nonuse of routine CMR over TTE for follow-up in patients receiving cytotoxic chemotherapy and cited that the incremental benefit of CMR over TTE does not justify the additional cost.