If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Corresponding author: Dr Brian Clarke, 477A-4th Fl, Burrard Bldg, St. Paul’s Hospital, 1081 Burrard St, Vancouver, British Columbia V6Z1Y6, Canada. Tel.: +1-604-875-5759; fax: +1-604-806-9927.
Pulmonary artery pressure (PAP) monitoring reduces heart failure (HF) hospitalizations (HFHs) and improves quality of life in New York Heart Association (NYHA) class III HF. We evaluated the impact of PAP monitoring on outcomes and health spending in a Canadian ambulatory HF cohort.
Methods
Twenty NYHA III HF patients underwent wireless PAP implantation at Foothills Medical Centre, Calgary, Alberta. Baseline, and 3-, 6-, 9-, and 12-month assessments of laboratory parameters, hemodynamics, 6-minute walk text and Kansas City Cardiomyopathy Questionnaire scores were collected. Healthcare costs 1 year pre- and post-implantation were collected from administrative databases.
Results
Mean age was 70.6 years; 45% were female. Results were as follows: an 88% reduction in emergency room visits (P = 0.0009); an 87% reduction in HFHs (P < 0.0003); a 29% reduction in heart function clinic visits (P = 0.033), and a 178% increase in nurse calls (P < 0.0002). Questionnaire and 6-minute walk test scores at baseline vs last follow-up were 45.4 vs 48.4 (P = 0.48) and 364.4 vs 402.8 m (P = 0.58), respectively. Mean PAP at baseline vs follow-up was 31.5 vs 24.8 mm Hg (P = 0.005). NYHA class improved by at least one class in 85% of patients. Mean measurable HF-related spending preimplantation was CAD$29,814 per patient per year and postimplantation was CAD$25,642 per patient per year (including device cost).
Conclusions
PAP monitoring demonstrated reductions in HFHs, and emergency room and heart function clinic visits, with improvements in NYHA class. Although further economic evaluation is needed, these results support the use of PAP monitoring as an effective and cost-neutral tool in HF management in appropriately selected patients in a publicly funded healthcare system.
Résumé
Contexte
La surveillance de la pression artérielle pulmonaire (PAP) réduit les hospitalisations liées à l’insuffisance cardiaque (HIC) et améliore la qualité de vie des patients atteints d’insuffisance cardiaque (IC) de classe III de la New York Heart Association (NYHA). Nous avons évalué l’effet de la surveillance de la PAP sur les résultats et les dépenses en santé dans une cohorte de patients ambulatoires atteints d’IC au Canada.
Méthodologie
Vingt patients atteints d’IC de classe III de la NYHA se sont fait implanter un dispositif sans fil de surveillance de la PAP au Foothills Medical Centre, à Calgary, en Alberta. Nous avons évalué les patients au départ et aux mois 3, 6, 9 et 12 en fonction des paramètres de laboratoire, de l’hémodynamique, des scores au test de marche de 6 minutes et au questionnaire de cardiomyopathie de Kansas City. Les coûts liés aux soins de santé un an avant et après l’implantation du dispositif ont été extraits de bases de données administratives.
Résultats
L’âge moyen des patients est de 70,6 ans et 45 % sont des femmes. Les résultats vont comme suit : une réduction de 88 % des visites à l’urgence (p = 0,0009); une réduction de 87 % des HIC (p < 0,0003); une réduction de 29 % des visites à une clinique de fonction cardiaque (p = 0,033) et une augmentation de 178 % des appels du personnel infirmier (p < 0,0002). Les scores initiaux au questionnaire et au test de marche de 6 minutes par rapport à ceux du dernier suivi sont respectivement de 45,4 contre 48,4 (p = 0,48) et de 364,4 contre 402,8 m (p = 0,58). La PAP moyenne au départ par rapport au suivi est de 31,5 contre 24,8 mmHg (p = 0,005). La classe de l’IC selon la NYHA s’est améliorée d’au moins une position chez 85 % des patients. Les dépenses mesurables moyennes liées à l’IC avant l’implantation sont de 29 814 $ CA par patient par année et de 25 642 $ CA par patient par année après l’implantation (incluent le coût du dispositif).
Conclusions
La surveillance de la PAP entraîne une réduction des HIC ainsi que des visites à l’urgence et à la clinique de fonction cardiaque, et une amélioration de la classe de l’IC selon la NYHA. Bien qu’une évaluation économique plus approfondie soit nécessaire, ces résultats militent en faveur de la surveillance de la PAP comme outil de prise en charge de l’IC sans incidence sur les coûts chez les patients adéquatement choisis dans un système de soins de santé financé par les fonds publics.
Heart failure (HF) is a prevalent and problematic disease and continues to have significant morbidity and mortality in the modern era despite expanded pharmacologic options, device-based therapies, and chronic disease management programs such as heart function clinics (HFCs).
According to government of Canada health surveillance data from 2012-2013, approximately 669,600 Canadians aged over 40 years were living with HF (3.6% of the population), and approximately 92,900 (5.2 per 1000) Canadians receive a diagnosis of HF each year.
In addition to the population impact, HF has significant health systems impacts and accounts for a significant portion of healthcare spending. Direct care costs of HF exceed USD$30 billion annually in the US, and are more than CAD$2.8 billion in Canada.
In 2013, the mean length of stay for an adult patient hospitalized for HF in Canada was 8.3 days, with a mean annual admission cost of CAD$10,970 per patient.
HF hospitalizations (HFHs) are associated with increased mortality and risk of subsequent all-cause and HF admissions, with approximately 25% of patients readmitted within 30 days of discharge.
Despite recent advances in HF drug and device treatments, a recent Canadian study of the Canadian Institute for Health Information (CIHI) hospital morbidity database reports no significant appreciable change in HFH prevalence, readmission rate, or length of stay between 2009 and 2018; these were, respectively, a prevalence of 216 per 100,000 population, a 20.6% national crude all-cause 30 day readmission rate ( > 50% due to HF), and a median length of stay of 7 days.
Routine methods for monitoring congestion (weight, symptoms, physical exam) are crude, unreliable, require frequent in-person assessments, and occur late in the pathway to HFH. Wireless PAP monitoring provides more-accurate assessment of left ventricular (LV) filling pressures and enables remote monitoring and management of HF patients. The fundamental principle in the approach to managing patients with wireless PAP monitoring is to use pharmacologic interventions aimed at lowering PAPs to normal, or to as near normal as possible. In most patients with HF, the diastolic PAP approximates the pulmonary capillary wedge pressure (LV filling pressure). The hemodynamic goals are to adjust pharmacotherapy as tolerated, to achieve target hemodynamics (where possible) as outlined in the, CardioMEMS Heart Sensor Allows Monitoring of Pressure to Improve Outcomes in NYHA Class III Patients (CHAMPION) trial
—that is, a target PAP as follows: systolic, 15-35 mm Hg; diastolic, 8-20 mm Hg; and mean, 10-25 mm Hg.
Achievement of these targets is attempted by means of using a combination of afterload reduction/vasodilators and diuretics (inclusive of sodium-glucose transporter-2 [SGLT-2} inhibitors) while monitoring the typical parameters monitored in HF patients—blood pressure, electrolytes, and renal function. These targets can be individualized to the patient based on clinical response, clinical stability, renal function, and blood pressure. The CardioMEMS system (Abbott Laboratories, Atlanta, GA) also records resting heart rate, which enables remote titration of beta-blockers once LV filling pressures are optimized. Once optimal status is achieved, subsequent medical interventions are aimed at maintaining stable pulmonary pressures using additional/adjunctive diuretics and/or vasodilators when PAPs begin to rise, signalling the beginning of a decompensation.
In 2011, the pivotal CHAMPION trial demonstrated that using the CardioMEMS wireless PAP monitor to guide outpatient HF management significantly reduced HFHs by 37% over 15 months, lowered the risk of death or first HF-related hospitalization, and improved quality of life; these effects occurred irrespective of LV ejection fraction (LVEF).
Since then, several other real-world and registry studies have demonstrated the effectiveness and safety of this device in other cohorts, routinely exceeding the benefits seen in the clinical trial.
Design and rationale of haemodynamic guidance with CardioMEMS in patients with a left ventricular assist device: the HEMO-VAD pilot study: PA guided LVAD management.
Pulmonary artery pressure-guided therapy in ambulatory patients with symptomatic heart failure: the CardioMEMS E uropean M onitoring S tudy for H eart F ailure (MEMS-HF ).
Lower rates of heart failure and all-cause hospitalizations during pulmonary artery pressure-guided therapy for ambulatory heart failure: one-year outcomes from the CardioMEMS post-approval study.
Remote hemodynamic monitoring equally reduces heart failure hospitalizations in women and men in clinical practice: a sex-specific analysis of the CardioMEMS post-approval study.
The CardioMEMS system was approved by the US Food and Drug Administration in 2014 for clinical use in New York Heart Association (NYHA) class III HF patients with a HFH within the preceding 12 months. Health Canada’s approval came in 2017 for the same indication, but uptake of wireless PAP-monitoring technology to assist in the management of HF patients in Canada has been delayed. In this study, we report clinical outcomes and the cost-utility of wireless PAP monitoring in a single-centre Canadian cohort of ambulatory NYHA class III HF patients with previous HFH in Calgary, Alberta, Canada.
Methods
Population
Between September 2018 and November 2020, a total of 21 adult patients (age > 18 years) with NYHA class III HF and at least one HFH provided consent for CardioMEMS implantation at Foothills Medical Centre in Calgary, Alberta. All patients in this study were enrolled from a heart function clinic (HFC) and received standard education, pharmacologic and nonpharmacologic counseling and management, including but not limited to fluid-intake monitoring, daily weights, blood pressure management and recording, and dietary and exercise recommendations. Over the following year, all patients received a combination of required in-person assessments at 3-month intervals for outcomes evaluation and remote telephone visits with management recommendations that incorporated PAP data.
Management of hemodynamic parameters
Patients implanted with the CardioMEMS device are instructed to take a reading daily, using the patient pillow they take home with them. They were also instructed to take routine home blood pressure readings. The device interrogation typically takes < 5 minutes, and patients are educated about how to do this on the day of implantation. When interrogated, the device transmits a snapshot of the systolic PAP, diastolic PAP, mean PAP, and heart rate via cellular signal or wireless networks to a secure server (merlin.net), and a visual scale for trends over time is plotted in addition to being provided in a daily numeric format (Fig. 1). The clinician uses this information to inform medication changes to optimize PAPs to the targets described in the first section of this paper. PAPs were checked twice per week during the first 3 months (optimization phase) by the monitoring physician. Medication changes are frequently made during the optimization phase to achieve the following: (i) improve and stabilize PAPs; and (ii) titrate guideline-directed medical therapy to target or maximal-tolerated doses in the case of patients with HF with reduced ejection fraction (HFrEF). Medication recommendations were communicated to the patients’ HFC nurse clinician to be communicated to the patient. Once PAPs are optimized (maintenance phase), the physician can set upper and lower thresholds that trigger an e-mail notification from merlin.net to be sent to the clinicians when these thresholds are crossed, thereby reducing the need for the physician to actively check the PAPs. The physician can adjust and individualize the thresholds specifically for each patient. In the maintenance phase, PAPs were checked twice per month, primarily by the HFC nurse clinician, and reviewed by the physician in situations of deviations from threshold. Depending on the patient, more- or less-frequent active PAP review may be needed.
Figure 1Typical web page display for a patient with a CardioMEMS device (Abbott Laboratories, Atlanta, GA), highlighting trends of pulmonary artery (PA) pressures over time. Red line = systolic PA pressure; green line = diastolic PA pressure; blue line = mean PA pressure; yellow lines show medication changes.
Elevated PAPs above a patient’s threshold are considered to be volume-overloaded, and diuretics and/or vasodilators were adjusted based on clinical judgement that considered the following: knowledge of hemodynamic profile from the implant right heart catheterization, and patients’ blood pressure, electrolyte levels, and renal function. Treatments could include the following: (i) increase or addition of loop diuretic; (ii) increase or addition of mineralocorticoid receptor antagonist; (iii) change to another loop diuretic; (iv) addition of adjunctive diuretic (ie, thiazide); (v) use of intravenous loop diuretic; and/or (vi) addition and/or increase in vasodilators. In cases of mild elevation in PAPs and satisfactory blood pressure, increasing the dose or number used of vasodilators (increasing afterload reduction) was often sufficient to reduce PAPs. Low PAPs below a patient’s threshold are considered to be volume-deplete, and treatment options included the following: (i) lower the dose of or discontinue diuretics; and/or (ii) consider liberalization of salt and fluid restriction. If low PAPs were associated with symptomatic hypotension, use of vasodilators was reduced or held. Electrolyte levels and renal function were re-evaluated with any change in medication, and PAPs were reassessed twice weekly until they were stabilized.
Clinical outcomes
Baseline, and 3-, 6-, 9- and 12-month assessments of n-terminal pro-B-type natriuretic peptide (NT-proBNP); creatinine; systolic, diastolic, and mean PAP; 6-minute walk test (6-MWT); and NYHA class were prospectively collected. Electronic medical records and paper charts were audited to determine the number of days spent in the hospital for HF during the year before and year after wireless PAP implantation. Electronic medical records were used to distinguish between days spent on general cardiology wards vs those spent on cardiac intensive care units. Similarly, the number of emergency room (ER) visits for HF, HFC medical doctor (MD) visits, HFC nurse clinician visits, and nurse clinician phone calls were recorded for the year before and year after device implantation.
Patient-reported outcomes
To obtain insight on the patient experience, the Kansas City Cardiomyopathy Questionnaire (KCCQ12) was prospectively collected at baseline and at 3-month intervals thereafter. Moreover, a second survey, created by our group (Appendix 1), was also administered at a single point in time after device implantation. This survey consisted of 6 questions scored on a 5-point Likert scale to determine overall satisfaction with our remote monitoring program. This survey was created by our group to capture the patient experience and has not been validated.
Appendix 1Patient reported outcomes survey
1.
Before receiving the CardioMEMS device, how confident were you in your heart failure care? Please circle the number that best describes your confidence.Scored from 1 (not confident) to 5 (very confident).
2.
After receiving the CardioMEMS device, how confident are you in your heart failure care?Scored from 1 (not confident) to 5 (very confident).
3.
Before receiving the CardioMEMS device, how worried were you about your heart failure?Scored from 1 (not worried) to 5 (very worried).
4.
After receiving the CardioMEMS device, how worried are you about your heart failure?Scored from 1 (not worried) to 5 (very worried).
5.
Patients that have the CardioMEMS device often require less in-person visits to their doctor. Do you prefer to have your heart health care provided remotely (via telephone using information from your device) or in-person?I strongly prefer remote visits.I slightly prefer remote visits.I prefer a mix of both.I slightly prefer in-person visits.I strongly prefer in-person visits.
6.
Based on your overall experience, how likely are you to recommend the CardioMEMS device to other people living with heart failure?Scored from 1 (not likely at all) to 5 (very likely).
7.
What is the most positive impact that the CardioMEMS device has had on your life?
8.
What is the most negative impact that the CardioMEMS device has had on your life?
Healthcare utilization costs for the 1 year pre-device implantation and 1 year post-device implantation were collected using the clinical outcomes obtained from medical records and Alberta Health Services (AHS) costs for services in the specific year they were delivered. All cost and pricing information is reported in CAD. We defined healthcare utilization crudely, as the combination of outpatient HF and inpatient HFH care. Outpatient HF care included physician billing associated with HFC visits, physician billing associated with ER visits, total overhead costs per HFC visit and ER visit, cost of device implantation, and estimation of billable monitoring costs over 1 year. The cost of device implantation was defined as the device cost, plus the cost of a day surgery visit and the physician billing cost of performing a right heart catheterization and pulmonary angiography. Inpatient costs were determined by summing the cost associated with the total number of days spent in a critical care unit with the total number of days spent on a medical ward. Monitoring costs were estimated by summing a physician limited follow-up billing code for the recommended PAP monitoring schedule, allowing for 2 deviations from thresholds per patient during the maintenance phase (CAD$25.09 per encounter to a total of CAD$1304.68 per patient per year).
We calculated the total number of days in hospital, ER visits, MD HFC visits, and nurse clinician HFC visits, multiplied by AHS cost data and physician billing codes associated with each of these encounters. The cost of device implantation was calculated as the sum of the cost of day surgery at an AHS facility, the cost of the device, and the physician billing associated with performing a right heart catheterization and pulmonary angiography. All costing information was obtained from AHS established rates and physician billing codes. Ambulatory (ER and outpatient), hospitalization, and procedure costs (same-day outpatient cardiac catheterization) were inclusive of staffing costs (nursing and administrative) as reported by AHS. Physician billing codes for ER, ambulatory, and in-hospital care were obtained from the Alberta Medical Association fee guide.
Statistical analysis
Descriptive statistics are presented as means with standard deviations. In comparing baseline and follow-up characteristics, normally distributed, continuous variables, including estimated glomerular filtration rate, PAPs, KCCQ12 scores, and 6-MWT scores, were assessed by Student t tests. Variables with skewed distribution, including NT-proBNP level, and all healthcare utilization and cost data, were compared with Wilcoxon signed-rank tests. All Student t tests were completed in Excel (Microsoft, Redmond, WA), and STATA (version 14.2, StataCorp, College Station, TX) was used to compute Wilcoxon signed-rank tests. Two-sided P values < 0.05 were considered significant. In the absence of follow-up data, the last observation carried forward approach was used to account for missing data.
Results
Of 21 patients, 19 were successfully monitored using the wireless PAP monitor. One patient had an unsuccessful implantation due to multiple retained right ventricular leads from an intracardiac device, and 1 patient had a successful implantation but developed pressure dampening due to a small implant artery 6 weeks after implantation. The 1 patient with pressure dampening was included in the primary analysis. However, a secondary analysis was performed with this patient excluded. No procedural complications occurred. The mean patient age was 70.6 ± 13.1 years, and 45% were female. HFrEF accounted for 90% of cases. Of the 20 patients followed, 11 had an ischemic cardiomyopathy (1 with HF with preserved ejection fraction [HFpEF]), 5 had a dilated cardiomyopathy, 1 had infiltrative disease (wild-type amyloid with HFrEF), 1 had idiopathic HFpEF, 1 had congenital heart disease with HFrEF, and 1 had a chemotherapy-induced cardiomyopathy (Table 1).
Table 1Baseline patient characteristics at the time of implantation of the CardioMEMS device (Abbott Laboratories, Atlanta, GA)
Variable
Total (n = 20)
Age, y
70.6 ± 13.1
Female sex
9 (45)
HFrEF
18 (90.0)
Diagnosis
Ischemic cardiomyopathy
11 (55)
Dilated cardiomyopathy
5 (25)
Infiltrative cardiomyopathy
1 (5)
Congenital heart disease
1 (5)
Chemotherapy-induced cardiomyopathy
1 (5)
Idiopathic HFpEF
1 (5)
Values are mean ± standard deviation, or n (%).
HFrEF, heart failure with reduced ejection fraction; HFpEF, heart failure with preserved ejection fraction.
Clinical measures at baseline and 12 months postimplantation are shown in Table 2. Mean estimated glomerular filtration rates at baseline and 12 months postimplantation were 49.0 ± 17.7 vs 45.5 ± 17.3 mL/min per 1.73 m2 (P = 0.54). Follow-up NT-proBNP values were skewed by 1 patient who had a NT-proBNP level of > 70,000 pg/mL at 12-months follow-up. The patient had wild-type transthyretin cardiac amyloidosis and had developed progressive ventricular dysfunction. After excluding the patient with a very high NT-proBNP level, a significant decline occurred in NT-proBNP level over time (mean: 2422 pg/mL ± 1729 at baseline vs 1462 pg/mL ± 1419 at 12 months, P = 0.038; Fig. 2A). Mean NT-proBNP levels at baseline and 12 months postimplantation, with the outlier included, were 2422 pg/mL ± 2467 vs 5270 pg/mL ± 16,213 (P = 0.47; Fig. 2B). We did not exclude this patient from any other aspect of this analysis. NT-proBNP level increased in some patients, and we suspect that this is due to a combination of baseline values being determined when patients had been stabilized after a recent hospitalization, progression of disease, cyclic levels of stability, and other factors that are known to affect NT-proBNP level (eg, renal function, atrial arrhythmias).
Table 2Clinical measures at baseline vs time of last-follow-up
Variable
Baseline
12 months postimplantation of the CardioMEMS device
P
eGFR, mL/min per 1.73 m2
49.0 ± 17.7
45.5 ± 17.3
0.54
NT-proBNP, pg/mL
2422 ± 2,466.7
5270 ± 16,213.2
0.47
NT-proBNP (outlier excluded)
2422 ± 1729
1462 ± 1419
0.038
Systolic PAP, mm Hg
46.9 ± 9.0
38.1 ± 9.0
0.006
Diastolic PAP, mm Hg
24.0 ± 5.8
18.7 ± 5.2
0.009
Mean PAP, mm Hg
31.5 ± 6.5
24.8 ± 6.7
0.005
KCCQ12 score
45.4 ± 13.8
48.4 ± 12.1
0.48
6-MWT, m
364.4 ± 169.6
402.8 ± 182.6
0.58
Values other than P are mean ± standard deviation.
eGFR, estimated glomerular filtration rate; KCCQ12, Kansas City Cardiomyopathy Questionnaire; NT proBNP, n-terminal pro-B-type natriuretic peptide; PAP, pulmonary artery pressure; 6-MWT, 6-minute walk test.
Figure 2(A) N-terminal pro-B-type natriuretic peptide (NT-proBNP) over time for each study patient following CardioMEMS device (Abbott Laboratories, Atlanta, GA) implantation. Black trend line represents mean NT-proBNP of the cohort (outlier excluded). (B) NT-proBNP over time for each study patient following CardioMEMS device implantation. Black trend line represents mean NT-proBNP of the cohort (outlier included).
Systolic, diastolic, and mean PAP at baseline and 12 months postimplantation were 46.9 mm Hg ± 9.0 vs 38.1 mm Hg ± 9.0 (P = 0.006), 24.0 mm Hg ± 5.8 vs 18.9 mm Hg ± 5.2 (P = 0.009), and 31.5 mm Hg ± 6.5 vs 24.8 mm Hg ± 6.7 (P = 0.005), respectively (Table 2; Fig. 3).
Figure 3Pulmonary artery pressure (PAP) over time for each study patient. (A) Pulmonary artery systolic pressure. (B) Pulmonary artery diastolic pressure. (C) Pulmonary artery mean pressure. Means are shown by black trend lines.
One death occurred, at 1 year plus 1 day postimplantation. This patient had sudden death during sleep, with no evidence of preceding worsening pulmonary pressures and no ventricular tachycardia or ventricular fibrillation detected from the implanted cardioverter defibrillator. KCCQ12 and 6-MWT scores at baseline and last follow-up were 45.4 ± 13.8 vs 48.4 ± 12.0 (P = 0.48) and 364.4 m ± 169.6 vs 402.8 m ± 182.6 (P = 0.58), respectively (Table 2). All patients were NYHA class III at the time of wireless PAP monitor implantation. At 12 months, the median NYHA class was II (Fig. 4), with 17 patients (85%) demonstrating improvement in their NYHA class.
Figure 4Percentage of patients by New York Heart Association (NYHA) class at baseline (Pre-MEMS) and 12-months after (post-MEMS) CardioMEMS device (Abbott Laboratories, Atlanta, GA) implantation.
Significant reductions occurred in number of ER visits (88%; P = 0.0009), absolute HFHs (87%; P < 0.003), physician HFC visits (29%; P = 0.033), and nurse HFC visits (28%; P = 0.033), with a 178% increase in nurse clinician phone calls (P < 0.0002; Table 3). In the entire cohort, a total of 229 days were spent in hospital due to HF in the year prior to implantation, and 31 days were spent in hospital due to HF in the year after implantation. The mean number of days in hospital per patient before and after wireless PAP monitor implantation were 11.4 ± 9.8 and 1.55 ± 5.7 days, respectively (P = 0.0003), corresponding to an 86% reduction in days in the hospital (Fig. 5). The mean number of admissions per patient before and after wireless PAP monitoring were 1.15 ± 0.75 and 0.15 ± 0.49, respectively (P = 0.0003), corresponding to an 87% reduction in number of admissions (Fig. 5). Results were similar, after excluding the patient with pressure dampening, for mean number of days spent in the hospital (11.2 ± 10.1 vs 0.3 ± 1.4 days, P < 0.001) and total number of hospitalizations (1.16 ± 0.8 vs 0.05 ± 0.23, P < 0.001).
Table 3Healthcare utilization preimplantation and postimplantation of the CardioMEMS device (Abbott Laboratories, Atlanta, GA) for patients who have completed 1 year of follow-up
Figure 5Mean absolute number of heart failure (HF) hospitalizations and days in hospital due to HF per patient over the 1 year preceding (Pre-MEMS) and 1 year following (Post-MEMS) CardioMEMS device (Abbott Laboratories, Atlanta, GA) implantation.
A total of 324 medication changes for 20 patients were made over the 1-year monitoring period. Diuretics were the most commonly changed medication class, with 138 changes made in loop diuretics. A total of 41 changes were made in beta-blockers, 53 in angiotensin antagonists, and 33 in aldosterone antagonists. The majority of medication changes were made within the first 3 months postimplantation (165 changes representing 50.9% of all medication changes; Fig. 6).
Figure 6Cumulative number of medication changes over the 12-month period following device implantation displayed as total changes per 3-month period for each drug class. ACEi, angiotensin-converting enzyme inhibitor; ARB, angiotensin receptor blocker; ARNi, angiotensin receptor-neprilysin inhibitor; BB, beta-blocker; MRA, mineralocorticoid receptor antagonist; SGLT2i, sodium-glucose transporter-2 inhibitor.
The mean cost of HFH in the year prior to, and after, wireless PAP monitor implantation was CAD$27,190.55 ± CAD$30,840.02 and CAD$4379.35 ± CAD$17,058.93 (P = 0.0009), respectively. The mean cost of healthcare utilization for HF in the year prior to implantation for all patients was CAD$29,813.62 ± CAD$30,780.65, whereas the total costs in the year after implantation (including and excluding the device cost) were CAD$25,642 ± CAD$17,276.21 and CAD$7184 ± CAD$17,276.21, respectively (Table 4). Results were similar after excluding the patient with pressure dampening for mean healthcare utilization for HF before (CAD$29 634 ± CAD$31,613) and after CardioMEMS implantation (including the device cost; CAD$23,207 ± CAD$2616, P = 0.046).
Table 4Healthcare utilization costs preimplantation and postimplantation of the CardioMEMS device (Abbott Laboratories, Atlanta, GA) for patients who have completed 1 year of follow-up
Type of healthcare utilization cost
Preimplantation (n = 20)
Postimplantation (n = 20)
P
HF clinic visits
1472 ± 1187
1059 ± 722
0.37
HF ER visits
637 ± 529
73 ± 240
0.034
Hospitalization (due to HF)
27,191 ± 30,840
4379 ± 17,059
0.0009
Total healthcare utilization due to HF (including device)
29,814 ± 30,781
25,642 ± 17,276
0.99
Total healthcare utilization due to HF (excluding device)
Regarding patient-reported outcomes, 73% of patients reported improvement in their confidence regarding their HF management (Fig. 7A), and 73% reported worrying less about living with HF (Fig. 7B). A total of 73% of patients also reported favoring either a mix of remote and in-person care or solely remote care. A total of 93% of patients were likely to recommend wireless PAP monitoring to a friend with HF.
Figure 7Sankey diagrams illustrating baseline and 1 year postimplantation of the CardioMEMS device (Abbott Laboratories, Atlanta, GA): (A) patient-reported confidence in heart failure management, on a 5-point Likert scale; (B) patient-reported worry regarding living with heart failure, on a 5-point Likert scale.
In this study, we describe the first reported cohort experience with remote PAP monitoring in the Canadian healthcare setting. All patients in our cohort were implanted based on Health Canada–approved clinical indications. Although approved for HF with any ejection fraction, most patients in our study had HFrEF, and the proportions of men and women were nearly equal.
Our study demonstrated that over a 1-year period, patients experienced significant reduction in PAP. We observed a 21% mean reduction in mean PAP at 12 months. This result is similar to results in other observational studies, which have reported reductions in mean PAP of 7%-17% after 12 months of therapy guided by wireless PAP monitoring.
Pulmonary artery pressure-guided therapy in ambulatory patients with symptomatic heart failure: the CardioMEMS E uropean M onitoring S tudy for H eart F ailure (MEMS-HF ).
Lower rates of heart failure and all-cause hospitalizations during pulmonary artery pressure-guided therapy for ambulatory heart failure: one-year outcomes from the CardioMEMS post-approval study.
Changes in PAP reported in the CHAMPION and the Hemodynamic-Guided Management of Heart Failure (GUIDE-HF) randomized-controlled trials were described by area-under-the-curve analysis; results of both trials were also consistent with reductions in PAP.
The decline in PAPs is directly related to the pharmacologic interventions aimed at lowering them. Up-titrating and/or adding afterload reducing agents/vasodilators (angiotensin-converting enzyme inhibitors, angiotensin receptor blocker, angiotensin receptor-neprilysin inhibitor, hydralazine/nitrates) can improve LV performance and lower LV filling pressures in HFrEF. Adjustment of diuretic dosing and/or addition of adjunctive diuretics (ie, mineralocorticoid receptor antagonists, SGLT-2 inhibitors, metolazone) is another important pharmacologic intervention to lower PAP in all types of HF and is the mainstay of targeted pharmacologic intervention in HFpEF, other than identifying and managing contributing comorbidities. Indeed, we observed significant medication changes, most commonly diuretics, and that > 50% of medication changes occurred in the first 3 months. Having access to real-time invasive hemodynamics also enables targeted and individualized pharmacologic intervention. We are able to see, in real time, the immediate hemodynamic effect of various medications on PAP. We are able to identify what decongestive strategy works for a particular patient. A patient may not respond to a higher-dose furosemide, but may respond to a switch to bumetanide or the addition of spironolactone or an SGLT-2 inhibitor. Monitoring PAP with the CardioMEMS device goes hand in hand with conventional physiologic monitoring of the other important factors that guide our management—namely, blood pressure and heart rate. We are able to identify which patients are optimally decongested, to focus on beta-blocker up-titration, and we receive resting heart rates with each patient transmission. With the reduction in PAP, 85% of our patients reported an improvement in NYHA class.
We also evaluated and report the effect on HFHs, ER visits, and HFC visits 1 year before and 1 year after device implantation. Despite the limitations associated with studying each patient as an internal control, the method of evaluating a period 1 year before and 1 year after CardioMEMS implantation has been used in at least 2 other studies.
Pulmonary artery pressure-guided therapy in ambulatory patients with symptomatic heart failure: the CardioMEMS E uropean M onitoring S tudy for H eart F ailure (MEMS-HF ).
Lower rates of heart failure and all-cause hospitalizations during pulmonary artery pressure-guided therapy for ambulatory heart failure: one-year outcomes from the CardioMEMS post-approval study.
In our study, a significant reduction in hospitalizations during the 1 year after implantation was observed. The average number of hospitalizations per patient in the 1 year before and 1 year after CardioMEMS implantation was 1.15 and 0.15, respectively, corresponding to a reduction of 87%. To our knowledge, this reduction in HFHs is the largest observed in managing patients with the use of the PAP data from the CardioMEMS device to date. The randomized controlled CHAMPION and more recent, GUIDE-HF trials reported 30% and 18% relative reductions in HFHs, respectively.
Both of these figures were substantially lower than the treatment effect reported in 3 large observational studies. In a cohort of 1200 patients, Shavelle et al. reported a 57% reduction in HFHs in the year following CardioMEMS implantation.
Lower rates of heart failure and all-cause hospitalizations during pulmonary artery pressure-guided therapy for ambulatory heart failure: one-year outcomes from the CardioMEMS post-approval study.
In a similar analysis, the CardioMEMS European Monitoring Study for Heart Failure (MEMS-HF) trial of 234 NYHA III HF patients with prior HFH in the preceding 12 months in Germany, The Netherlands, and Ireland, reported a 62% reduction in HFHs over the following 1 year.
Pulmonary artery pressure-guided therapy in ambulatory patients with symptomatic heart failure: the CardioMEMS E uropean M onitoring S tudy for H eart F ailure (MEMS-HF ).
In the most recently published CardioMEMS HF System OUS Post Market Study (COAST)-UK observational study of 100 patients with NYHA class III HF and HFH in the preceding 12 months in the United Kingdom, an 82% reduction in annualized HFH rates in the 1 year postimplantation, compared to rates in the 1 year prior to implantation, was observed.
Our population was similar to other US, European, and United Kingdom reports in terms of age, gender, etiology of HF, and renal function; however, our cohort was predominantly patients with HFrEF. Monitoring frequency was weekly in the MEMS-HF study. In the COAST-UK study, active monitoring frequency was at least 2-3 times per week until patients were stable, and weekly thereafter. In our study, we reviewed PAP data 2-3 times per week until optimal or best attainable PAPs were obtained (twice per week for 3 months), and twice per month thereafter. Therefore, once patients are optimized, active monitoring can be reduced without losing the clinical benefit. We similarly demonstrate that a significant number of medication changes are made and that the majority of these medication changes are performed in the first 3 months postimplantation.
The consistently superior performance of wireless PAP monitoring in observational studies, compared to that in randomized controlled trials, is notable. Perhaps, owing to their involvement in a clinical trial, the additional care received by patients in the control arms of the CHAMPION and GUIDE-HF trials attenuated the events they would have otherwise experienced. This possibility is supported by the observation that the HFH rate in the control arm of GUIDE-HF trial (0.497 events per patient-year) was markedly different from the preimplantation HFH rate in our study (1.15 events per patient-year) and that reported in other real-world studies (1.25-1.55 events per patient-year).
Pulmonary artery pressure-guided therapy in ambulatory patients with symptomatic heart failure: the CardioMEMS E uropean M onitoring S tudy for H eart F ailure (MEMS-HF ).
Lower rates of heart failure and all-cause hospitalizations during pulmonary artery pressure-guided therapy for ambulatory heart failure: one-year outcomes from the CardioMEMS post-approval study.
In interpreting these discrepancies, a notable finding is that even control patients in the CHAMPION and GUIDE-HF trials received the CardioMEMS device and that regular, scripted phone calls in which treatment advice was provided were incorporated as part of blinding. Perhaps the regularity of this advice attenuated HFHs in control patients, and the frequency of this follow-up may be challenging to achieve outside of a clinical trial setting. An additional consideration in our cohort is the small sample size.
Although data for mortality benefit associated with wireless PAP monitoring have not been sufficiently robust/convincing, we do know that HFHs carry prognostic significance, and patients with 1 HFH are at increased risk of death. A contemporary analysis of the Prospective Comparison of ARNi With ACEi to Determine Impact on Global Mortality and Morbidity in Heart Failure (PARADIGM-HF) trial demonstrated that any decompensation event (ER visit for HF, or HFH) increased the risk of death 5-fold.
Importance of clinical worsening of heart failure treated in the outpatient setting: evidence from the Prospective Comparison of ARNI with ACEI to Determine Impact on Global Mortality and Morbidity in Heart Failure Trial (PARADIGM-HF).
We also know that HFHs have an additive effect on overall mortality, with higher mortality and a lower median survival rate with each subsequent hospitalization.
During the 1-year postimplantation period, a rise in nurse clinician phone calls through our HFC was observed. However, in our experience, these phone calls were incorporated efficiently into clinic workflow; the calls, though short, were more frequent, to communicate medication changes and as a symptom check associated with the same. The increase in nurse clinician phone calls most commonly occurred during the first 3 months postimplantation (46% of all phone calls), owing to the frequent medication changes made in efforts to optimize PAPs (optimization phase). The increase was offset by significant reductions in ER visits for HF and outpatient MD visits, which were 88% and 29%, respectively, to produce a net effect of lowering healthcare utilization. That being said, additional nursing resources are expected to be required to manage the higher frequency of patient contact in larger monitoring programs. The anticipated global cost-savings in HFH reduction can be expected to offset the cost of this additional resource use. A noteworthy point is that MD visits were mandated every 3 months to collect outcome data; therefore, the number of MD visits that were medically necessary is likely much lower, and the reduction in MD visits is likely much higher than that reported here. Additionally, the COVID-19 pandemic affected our mandated every–3 month in-person follow-up visits, with many in-person visits being skipped due to the pandemic.
With respect to cost, we report an 80% reduction in healthcare utilization costs for HFHs, HFC visits, and ER visits for HF in the year following wireless PAP-monitor implantation. When the cost of the device implantation was added to the healthcare utilization cost postimplantation, a mean absolute spending reduction of CAD$4172 per patient-year was observed, as compared to the healthcare utilization cost in the year prior to implantation. However, even greater cost-savings were seen after exclusion of the 1 patient who had pressure dampening, which precluded clinical use of the sensor. Important to note is that we did not account for the cost of nurse clinician calls; however, as noted earlier, the calls were predominantly brief calls to communicate medication changes. The increased use of resources required for larger monitoring programs would be offset by a reduction in hospitalization costs.
These results are in contrast to those of Sandhu et al, who describe an increase in absolute cost associated with PAP monitoring.
In their study, published CHAMPION trial data over 30 months, and Medicare reimbursement data, were used to estimate costs of the CHAMPION trial treatment arm and control group. The treatment group was found to incur a cost of USD$176,648 per patient, whereas the control group incurred a cost of USD$156,569 per patient. Reasons for the discrepancy between their result and ours may include the use of different methodologies to determine cost. Although Sandhu et al. used a Markov model to estimate spending, we audited medical records to determine healthcare spending directly for each patient individually. Our method of using direct healthcare costs likely provides a more accurate reported cost. Our study is also based on Canadian data, whereas theirs is based on data from the US, where hospitalization costs differ substantially.
Although the study by Sandhu et al. reported an increase in absolute cost, they also reported that the device was cost effective, with an associated incremental cost-effectiveness ratio (ICER) of USD$71,462 per quality adjusted life-year (QALY) and USD$48,054 per life-year.
Pulmonary artery pressure-guided heart failure management: US cost-effectiveness analyses using the results of the CHAMPION clinical trial: cost-effectiveness of haemodynamic-guided heart failure care.
These results are comparable to the ICER of transcatheter aortic valve replacement, which has been reported to be USD$32,170 per QALY, compared to medical therapy over a 3-year time horizon.
Cost effectiveness of transcatheter aortic valve replacement compared to medical management in inoperable patients with severe aortic stenosis: Canadian analysis based on the PARTNER trial cohort B findings.
Although the results of our study are encouraging, they do have limitations. Important to note is that our study is limited by use of an internal control. When the year prior to implantation is used as a reference, as in our study, progression of HF may blunt the treatment impact of the CardioMEMS device. Having a propensity-matched, local control who did not receive therapy guided by PAP monitoring would likely have provided results more reflective of the true cost-benefit. To demonstrate HF stability prior to implantation, the 1-year period prior to implantation could have been divided into 2 intervals of 6 months each. We modelled our reporting on that in other published studies. Our study is also limited by a small sample size, which may have impacted the significance of our results. For example, we did not observe a significant change in mean KCCQ12 scores from baseline to 12 months, whereas in a larger study, a significant improvement in KCCQ12 score that was directly correlated to reductions in PAP has been reported.
Pulmonary artery pressure-guided therapy in ambulatory patients with symptomatic heart failure: the CardioMEMS E uropean M onitoring S tudy for H eart F ailure (MEMS-HF ).
We also do not report how many patients achieved the hemodynamic targets.
Completing this study during the COVID-19 pandemic also may have influenced patient healthcare utilization. The government of Alberta declared a state of public health emergency on March 17, 2020. Six patients were implanted following this date. In the GUIDE-HF trial, control arm outcomes appeared to fall during the COVID-19 pandemic, thus blunting the treatment effect.
Outcomes during the control period in the 6 patients who underwent device implantation following March 17, 2020 could have been attenuated in a similar fashion. Both the MEMS-HF and COAST-UK studies were reported in populations assessed prior to the COVID-19 pandemic.
Follow-up in this study was also limited by constraints of the COVID-19 pandemic. For example, of 100 planned 6-MWTs, only 36 were performed. A phenomenon of substantially reduced numbers of 6-MWTs during the COVID-19 pandemic was also observed in the GUIDE-HF trial.
Finally, all patients in our study met the Health Canada indication for wireless PAP monitoring at the time of implantation—that is, NYHA III symptoms with at least 1 HFH in the preceding year. The recently published GUIDE-HF trial expanded its inclusion criteria to include patients with NYHA II or IV heart failure, and those without recent admission but with an elevated NT pro-BNP level. The findings of our study cannot be applied to support the use of wireless PAP monitoring in these populations.
Conclusion
In this single-centre Canadian real-world experience study, use of wireless PAP monitoring with the CardioMEMS device demonstrated significant reductions in PAP, improvement in NYHA functional class, and a significant reduction in HFHs. Despite the upfront cost of the device, average spending was less in the year following device implantation, although not by a statistically significant margin. Although further economic evaluation is warranted, these results support the use of ambulatory hemodynamic monitoring in carefully selected patients to enhance outpatient HF management in a publicly funded healthcare system.
Funding Sources
The authors have no funding sources to declare.
Disclosures
R.M. reports receiving speaker honoraria and research support from Pfizer. G.S. reports receiving speaker honoraria and research support from Boston Scientific, and Medtronic. B.C. reports receiving speaker honoraria, consulting fees, and research support from Abbott. The other authors have no conflicts of interest to disclose.
References
Go A.S.
Mozaffarian D.
Roger V.L.
et al.
Heart disease and stroke statistics—2013 update: a report from the American Heart Association.
Design and rationale of haemodynamic guidance with CardioMEMS in patients with a left ventricular assist device: the HEMO-VAD pilot study: PA guided LVAD management.
Pulmonary artery pressure-guided therapy in ambulatory patients with symptomatic heart failure: the CardioMEMS E uropean M onitoring S tudy for H eart F ailure (MEMS-HF ).
Lower rates of heart failure and all-cause hospitalizations during pulmonary artery pressure-guided therapy for ambulatory heart failure: one-year outcomes from the CardioMEMS post-approval study.
Remote hemodynamic monitoring equally reduces heart failure hospitalizations in women and men in clinical practice: a sex-specific analysis of the CardioMEMS post-approval study.
Importance of clinical worsening of heart failure treated in the outpatient setting: evidence from the Prospective Comparison of ARNI with ACEI to Determine Impact on Global Mortality and Morbidity in Heart Failure Trial (PARADIGM-HF).
Pulmonary artery pressure-guided heart failure management: US cost-effectiveness analyses using the results of the CHAMPION clinical trial: cost-effectiveness of haemodynamic-guided heart failure care.
Cost effectiveness of transcatheter aortic valve replacement compared to medical management in inoperable patients with severe aortic stenosis: Canadian analysis based on the PARTNER trial cohort B findings.
00294-3&id=gr1.jpg){kind=link}
00294-3&id=gr2.jpg){kind=link}
00294-3&id=gr3.jpg){kind=link}
00294-3&id=gr4.jpg){kind=link}
00294-3&id=gr5.jpg){kind=link}
00294-3&id=gr6.jpg){kind=link}
00294-3&id=gr7.jpg){kind=link}