Time Spent Engaging in Health Care Among Patients With Left Ventricular Assist Devices

Abstract

Objectives:

This study aims to examine a novel patient-centered metric of time spent engaging in left ventricular assist device (LVAD)-related clinical care outside the home.

Background:

Although LVAD implantation can improve survival and functional capacity in patients with advanced heart failure, this may occur at the expense of significant time spent engaging in LVAD-related health care activities.

Methods:

The authors retrospectively assessed consecutive patients at a single center who received a continuous-flow LVAD between May 9, 2008, and December 31, 2019, and queried health care encounters after implantation, including all inpatient encounters and LVAD-related ambulatory encounters. Patient-level time metrics were determined, including the total number of days with any health care encounter, and the total estimated time spent receiving care. The primary outcome was the proportion (%) of days alive with an LVAD spent engaged in at least 1 health care encounter. The secondary outcome was the proportion (%) of total time alive with an LVAD spent receiving care.

Results:

Among 373 patients, the median number of days alive with LVAD was 390 (IQR: 158–840 days). Patients had a median number of 88 (IQR: 45–161) days with ≥1 health care encounter, accounting for 23.2% (IQR: 16.3%–32.4%) of their days alive with an LVAD. A median 6.0% (IQR: 2.1%–14.1%) and 15.0% (IQR: 10.7%–20.0%) of total days alive were spent in inpatient and ambulatory encounters, respectively. Patients spent a median of 592 (IQR: 197–1,257) hours receiving care, accounting for 5.6% (IQR: 2.2%–12.7%) of their total time alive with an LVAD.

Conclusions:

LVAD patients spent more than 1 of every 5 days engaging in health care. Our findings may inform strategies to improve efficiency of postdischarge care delivery and expectations for post-treatment care.

Left ventricular assist devices (LVADs) have emerged as an established therapy for select patients with advanced heart failure (HF). Because of enhanced patient selection, pump design, and post-implantation management, LVAD patients benefit from improved longevity, with 1-year survival rate >80%, as well as enhanced health-related quality of life (HRQOL).^1–4 Thus, quality improvement in LVAD care has expanded to include efforts to increase cost effectiveness, reduce resource usage, and decrease care burden for LVAD patients and caregivers. Studies to date have shown that LVAD therapy is associated with a high frequency of emergency department visits and hospital readmissions, and a high cost of care.^5–9 However, a key patient-centered variable—the amount of time spent engaging in health care after LVAD—has not previously been examined.

The management of chronic conditions is time-consuming and burdensome for patients and caregivers.^10–12 Patients with chronic conditions such as diabetes, chronic lung disease, and HF may spend more than 2 hours per day on health care–related activities.^13,14 With increasing disease complexity, the workload increases: 1 study found that patients with 3 chronic conditions would need to visit a health professional 1.2 to 5.9 times a month and spend a mean of 50 to 70 hours each month on health-related activities to meet consensus recommendations for chronic disease care.¹⁵ This work that patients do to manage their health is often underappreciated by clinicians, but has important implications for quality of life and adherence to treatment plans.^16–18 In addition, the time burden of therapy, or the amount of time that patients spend on health care–related activities, represents time that cannot be spent doing usual activities including household, leisure, workplace, and social pursuits.

The time patients with LVADs spend engaging in health care is likely substantial; however, it is understudied. Health care for LVAD patients care can take place inside the home (eg, dressing change, managing the device, and organizing and taking medications) or outside the home (eg, clinic visits, blood draws and diagnostic procedures, and days spent admitted to the hospital). A better appreciation of patient experiences with health care engagement after LVAD implantation may: 1) improve the process of shared decision-making before LVAD implantation; 2) establish expectations for self-care; and 3) inform targeted interventions to improve efficiency of post-LVAD care delivery. Thus, the purpose of this study is to describe time spent engaging in health care for LVAD-related care outside the home.

METHODS

Patient population.

We retrospectively assessed consecutive adult patients (age ≥18 years) who underwent durable, continuous-flow LVAD implantation at Northwestern Memorial Hospital, Chicago, Illinois, USA, between May 9, 2008, and December 31, 2019. We excluded patients who did not survive to discharge during the index hospitalization for LVAD surgery. Patients with right and biventricular assist devices were also excluded. This project was approved by the Northwestern University Institutional Review Board.

Sample characteristics.

Baseline clinical and demographic information and censor dates were abstracted from the Northwestern Mechanical Circulatory Support REDCap database. Study entry was the date of discharge from the index hospitalization for LVAD surgery. Censor date was the earliest of the following: heart transplantation, LVAD explant, transfer of care, death, or end of study (December 31, 2019). Subject study period (from study entry to censor) is termed “days alive with an LVAD.”

Health care encounters.

For each eligible patient, all health care encounters were queried from the electronic health record (EHR) by using the Northwestern Medicine Enterprise Data Warehouse (Northwestern University Clinical and Translational Sciences Institute, Chicago, Illinois, USA), which has stored clinical observations of >6.6 million patients since 2000.¹⁹ Specifically, we included all inpatient encounters (including hospital admissions and emergency department visits) and ambulatory encounters (including clinic visits, blood draws, and diagnostic tests and procedures) that were considered related to LVAD care or cardiovascular disease and associated comorbidities based on consensus by the study team, institutional practice guidelines, and American Heart Association and International Society for Heart and Lung Transplantation guideline recommendations.^20,21 Bloodwork collected outside of our health system (ie, at a local laboratory or clinic) is also entered into our health record for LVAD patients; therefore, these data were included in our query. Table 1 describes the health care encounters that were queried through the Northwestern Medicine Enterprise Data Warehouse and how they were defined. For reference, our Mechanical Circulatory Support Program Procedures, which detail our post-LVAD clinic and lab protocols, are provided in the Supplemental Appendix.

TABLE 1.

Data Elements Queried and Associated Time Estimates^a

Type of Encounter	Description	Time Spent
Inpatient	All inpatient admissions	Calculated based on check-in and check-out times
Emergency	Emergency department visits that did not result in an inpatient admission	Calculated based on check-in and check-out times
Ambulatory
Observation	Outpatient procedures during which patient was listed as being under observation status (ie, right and left heart catheterization)	Calculated based on check-in and check-out times
Clinic visit	Includes any ambulatory clinic visit with the following divisions/departments: cardiology, gastroenterology, endocrinology, neurology, infectious diseases, nephrology, palliative medicine, dermatology, general internal medicine/primary care, cardiothoracic surgery, general surgery, neurosurgery, and ear nose and throat	Estimated based on expert consensus and national survey data for visit duration (60 min)
Bloodwork	Includes an outpatient blood draw for 1 or more of the following laboratory tests: hemoglobin, creatinine, INR, aspartate aminotransferase, brain natriuretic peptide, lactate dehydrogenase, and thyroid stimulating hormone	Estimated based on expert consensus for visit duration (15 min)
Diagnostic tests and procedures	Includes the following tests and procedures: ECG, echocardiogram, PFTs, x-ray chest, x-ray abdomen, ultrasound, DEXA, endoscopies (colonoscopy, video capsule endoscopy, EGD), PICC placement, blood transfusion, and cardiac rehabilitation	Estimated based on expert consensus for visit duration, min  ECG: 15  Echocardiogram: 60  CT scan: 60  PFTs: 60  X-ray chest/abdomen: 30  Ultrasound: 60  DEXA: 60  Endoscopy: 60  PICC: 60  Transfusion: 120  Cardiac rehabilitation: 90

^aIdentified by Current Procedural Terminology codes.

CT = computed tomography; DEXA = dual-energy x-ray absorptiometry; ECG = electrocardiogram; EGD = esophagogastroduodenoscopy; INR = international normalized ratio; PFT = pulmonary function test; PICC = peripherally inserted central catheter.

Time estimates for health care encounters.

The total amount of time spent receiving care was calculated or estimated. For emergency department visits, hospital admissions, and some ambulatory encounters (eg, right heart catheterization), time spent on each health care encounter was calculated as the difference between the check-in and check-out times as recorded in the EHR. For all other encounters, visit timestamps are unreliable within the EHR system. Therefore, based on national survey data (National Ambulatory Medical Care Survey) and consensus among study team members (physicians, nurses, and LVAD nurse coordinators), estimated lengths of time for visits, tests, and procedures were generated to account for appointment and wait time for each encounter.²² Time estimates are listed in Table 1.

Quality check.

A manual review of EHR data was performed to assess the accuracy of the data query. Thirty patients (8%) were randomly selected from our cohort and adjudicated by manual review of events documented in the medical record. A total of 3,238 unique health care encounters from the 30 patients in the data query were reviewed and compared with the medical record. We identified 140 unique health care encounters in the medical record that were missing from the data query (4.3% missing encounters) and 86 health care encounters that appeared in the data query but did not appear in the medical record (2.0% extraneous encounters). Missing or extraneous encounters was mostly related to ambulatory bloodwork performed at other institutions (62%) or ambulatory imaging procedures (14%). Based on these rates and the nature of inaccurate encounters, the degree of missingness was considered acceptable.

Statistical analysis.

For each patient, the following time metrics were determined: days alive with an LVAD, number of unique health care encounter days (any day with at least 1 encounter), and total estimated time (hours) spent receiving care. The primary outcome was the proportion (%) of total days alive with an LVAD spent engaged in at least one health care encounter. The secondary outcome was the estimated proportion (%) of total time alive with an LVAD spent receiving care. Formulas describing the primary and secondary outcome are depicted in Figure 1.

FIGURE 1.

Primary and Secondary Outcomes

Definitions of primary and secondary outcomes are described. LVAD = left ventricular assist device.

To explore differential patterns of health care engagement partitioned by days alive with an LVAD, we identified patients who survived >1 year and determined the proportion of days spent in health care during the first year after LVAD implantation and then after the first year. The Wilcoxon signed sum rank test was used to compare proportion of days spent in health care during the first year and after the first year for the entire cohort and then stratified by implantation strategy (bridge-to-transplantation [BTT] or destination therapy [DT]). To explore differential patterns of health care engagement over time (eg, to determine whether changes over time and technology were associated with different levels of health care engagement), we identified patients who survived >1 year post-implantation and evaluated the proportion of total days alive with an LVAD spent engaged in at least 1 health care encounter during their first year after implantation. Using implantation year as a categorical predictor, we evaluated differences in proportion of days spent in health care through univariable linear regression.

Finally, subgroup analyses were performed. The primary and secondary analyses were repeated by age (≥ or <65 years), race (white, black, other), insurance type (commercial, Medicare, Medicaid), implantation strategy (BTT, DT), and device type (HeartMate II [St Jude Corp], HeartWare HVAD [HeartWare Corp], HeartMate III [Abbott]). The Kruskal-Wallis test was used to assess differences between subgroups.

Statistical significance was declared at 2-sided 5% alpha level. No adjustments for multiplicity were made. All analyses were performed using SAS Enterprise Guide version 8.1 (SAS Institute).

RESULTS

The cohort consisted of 373 patients. Baseline characteristics are listed in Table 2. Mean age was 55 ± 14 years, 79% were male, 54% were white, and 43% used Medicare as their primary insurance payer. Etiology of HF was nonischemic in the majority of patients (61%). Heartware HVAD was the most common implanted device (53%) followed by Heartmate II (34%), and Heartmate III (13%). Median days alive with LVAD was 390 (IQR: 158–840) days, with BTT was 272 (IQR: 133–599) days, and with DT was 521 (IQR: 245–1,016) days. During follow-up, 42% of patients underwent heart transplantation and 21% died. A minority of patients transferred care (6%) and underwent LVAD explantation (2%). The remaining patients were followed through the end of study period (29%).

TABLE 2.

Patient Characteristics (N = 373)

Age at implantation, y	55 ± 14
Sex
Male	293 (79)
Female	80 (21)
Race
Latino, Hispanic, or Spanish ethnicity	43 (11)
White	200 (54)
Black	96 (26)
Asian	16 (4)
American Indian/Alaska Native	3 (1)
Other or unknown	58 (15)
Comorbidities at implantation
Diabetes	134 (36)
Dyslipidemia	195 (52)
Hypertension	200 (54)
Chronic kidney disease	146 (39)
Atrial arrhythmias	180 (48)
Stroke	39 (10)
Peripheral vascular disease	26 (7)
Chronic obstructive pulmonary disease	66 (18)
Payer
Commercial	124 (33)
Medicare	162 (43)
Medicaid	51 (14)
Other or unknown	36 (10)
Marital status
Unmarried or single	94 (25)
Married or partnered	230 (61)
Unknown	49 (14)
HF etiology
Nonischemic cardiomyopathy	228 (61)
Ischemic cardiomyopathy	145 (39)
LVAD type
Heartmate II	127 (34)
Heartware HVAD	196 (53)
Heartmate III	50 (13)
Implant strategy
Bridge-to-transplantation	182 (49)
Destination therapy	191 (51)
INTERMACS profile
1	70 (19)
2	137 (37)
3	91 (24)
4	27 (7)
5–7	5 (1)
Unknown	43 (12)
Reason for censor
Transplantation	155 (42)
End of study period	111 (29)
Death	77 (21)
Transfer of care	23 (6)
Explantation	7 (2)

Values are mean ± SD or n (%).

HF = heart failure; INTERMACS = Interagency Registry for Mechanically Assisted Circulatory Support; LVAD = left ventricular assist device.

Table 3 shows summary estimates of days spent engaging in health care and time spent receiving care. Patients had a median of 88 (IQR: 45–161) unique health care encounter days, accounting for 23.2% (IQR: 16.3%–32.4%) of their total days alive with an LVAD. Days with inpatient encounters accounted for 6.0% (IQR: 2.1%–14.1%) of total days alive, whereas days with ambulatory encounters accounted for 15.0% (IQR: 10.7%–20.0%) of total days (Central Illustration). Patients spent a median of 592 (IQR: 197–1257) hours receiving care, accounting for 5.6% (IQR: 2.2%–12.7%) of patients’ total time alive with an LVAD. Inpatient care accounted for 4.6% (IQR: 1.9%–12.4%) of total time alive, whereas ambulatory care accounted for 0.3% (IQR: 0.2%–0.6%) of total time. The distribution of days alive with an LVAD and of the proportion (%) of total days alive with LVAD spent engaged in health care for the cohort are presented in Figure 2.

TABLE 3.

Survival Time Analysis

Median unique health care encounters, days	88 (45–161)
Inpatient	25 (7–53)
Ambulatory	54 (23–111)
Median health care encounter days as proportion of days alive with LVAD, %	23.2 (16.3–32.4)
Inpatient	6.0 (2.1–14.1)
Ambulatory	15.0 (10.7–20.0)
Median time receiving care, hours	592 (197–1,257)
Inpatient	496 (131–1,141)
Ambulatory	31.5 (9.5–80.0)
Median time receiving care as proportion of time alive with LVAD, %	5.6 (2.2–12.7)
Inpatient	4.6 (1.9–12.4)
Ambulatory	0.3 (0.2–0.6)

Values are median (IQR).

Abbreviation as in Table 2.

CENTRAL ILLUSTRATION:

Time Spent Engaging in Health Care Among Patients With Left Ventricular Assist Devices

Consecutive patients at a single center who received a continuous-flow left ventricular assist devices (LVADs) were retrospectively assessed and health care encounters after implantation were queried. The primary outcome was the proportion (%) of days alive with an LVAD spent engaged in at least 1 health care encounter. Patients spent median 23% of days alive with an LVAD engaging in at least 1 health care encounter. A median 6.0% and 15.0% of total days alive were spent in inpatient and ambulatory encounters, respectively.

FIGURE 2.

Distribution of Days Alive With an LVAD and Proportion of Days Spent Engaging in Health Care

(A) The distribution of days alive with an LVAD among the study cohort. (B) The proportion of total days alive with an LVAD spent engaged in health care. Abbreviation as in Figure 1.

There were 34,190 total health care encounters queried for this analysis. Fifty-nine percent of total encounters were blood draws, whereas 23.3% were clinic visits, and 5.8% were inpatient visits. The majority (66.2%) of clinic visits for the total cohort were in cardiology and cardiology-related subspecialties (eg, advanced HF and electrophysiology), followed by infectious diseases (11%), and cardiothoracic surgery (9.4%) (Figure 3).

FIGURE 3.

Proportion of Total Encounters in Each Encounter Type and Clinic Type

(A) The 34,190 total health care encounters queried for this analysis were stratified by encounter type. (B) Clinic visits were then stratified by specialty. CT = computed tomography; DEXA = dual energy x-ray absorptiometry.

There were 190 patients (51%) who survived >1 year after LVAD discharge (n = 72 BTT, n = 118 DT). Median days alive with an LVAD among these patients was 829 (IQR: 529–1,366) days for the total cohort, 738 (IQR: 513–1,392) days for BTT patients, and 853 (IQR: 584–1,309) days for DT patients. Proportion of days spent in health care during and after the first year is depicted in Figure 4. The total cohort and DT LVAD patients spent a lower proportion of total days in health care after the first year than during the first year; there was no significant difference in proportion of days engaged in health care during versus after the first year for BTT patients. There was no significant difference in proportion of total days alive with LVAD spent in inpatient encounters during versus after the first year with an LVAD for the total cohort, BTT patients, and DT patients. In contrast, LVAD patients who survived >1 year spent a smaller proportion of their total days alive in ambulatory care after the first year versus during the first year.

FIGURE 4.

Proportion of Days Alive With an LVAD Spent Engaging in Health Care Among Patients Who Survived >1 Year

Bar charts showing proportion of days alive with an LVAD spent engaging in health care among the total cohort (left), bridge-to-transplantation patients (center), and destination therapy patients (right) both during (blue bars) and after (orange bars) the first year post-implantation.*P < 0.05 by Wilcoxon signed rank test.

Abbreviation as in Figure 1.

The longitudinal analysis is represented in Figure 5. Between 2008 and 2018, there were significant differences over time in the proportion of days alive with an LVAD spent engaging in health care during the year after implantation (P = 0.0009).

FIGURE 5.

Longitudinal Analysis

Box and whisker plot showing proportion of days spent engaging in health care among patients who survived at least 1 year, by year of implantation. There was a strong association between year of implantation and proportion of days spent engaging in health care (P = 0.0009).

Subgroup analyses evaluating the impact of clinical, demographic, and device-related factors on the time metrics are listed in Table 4. The proportion of days alive with an LVAD spent engaging in health care encounters was higher for HVAD patients versus those with other devices. Additionally, there was a trend towards higher proportion of days spent engaging in health care for BTT versus DT patients. Otherwise, the proportion of unique health care days and encounter time was similar across all other subgroups.

TABLE 4.

Subgroup Analyses

	Proportion (%) of Days Alive With LVAD Spent Engaged in ≥1 Health Care Encounter	Proportion (%) of Total Time Alive With an LVAD Spent Receiving Care
Age, y
<65 (n = 262)	23.6 (16.3–33.1)	5.2 (2.1–12.5)
≥65 (n = 111)	22.1 (15.9–30.2)	5.9 (2.9–14.8)
P value	0.70	0.12
Race
White (n = 200)	22.2 (16.6–30.3)	5.2 (2.3–12.6)
Black (n = 96)	24.1 (14.3–32.5)	6.6 (2.1–14.8)
Other (n = 77)	24.9 (16.3–33.5)	5.1 (2.1–11.6)
P value	0.7	0.4
Implant strategy
Bridge to transplant (n = 182)	24.4 (17.3–33.1)	5.2 (2.1–12.3)
Destination Therapy (n = 191)	21.5 (14.9–31.8)	5.8 (2.3–13.9)
P value	0.05	0.45
Device type
HeartMate II (n = 127)	21.5 (14.2–31.6)	5.1 (2.1–10.2)
HeartWare HVAD (n = 196)	24.5 (17.4–33.9)	6.4 (2.5–13.5)
HeartMate III (n = 50)	21.8 (13.8–28.4)	5.3 (1.5–15.0)
P value	0.04	0.35

Values are median (IQR), unless otherwise indicated.

Abbreviation as in Table 2.

DISCUSSION

Our study is the first to explore a novel, patient-centered metric defined by time spent engaging in health care among LVAD patients. We observed that LVAD patients spent a considerable proportion of their days—more than 1 in every 5—engaging in some form of health care after discharge from LVAD implantation. The time spent receiving care also represented an estimated >5% of their total survival time. These findings were generally consistent across subgroups (by age, race, insurance status, and LVAD implantation strategy), with minor differences based on device type. In addition, although the proportion of days with health care encounters did decrease after 1 year of therapy among patients who survived at least 1 year, the overall proportion of days and time spent receiving care remained high. Further, over time and with improved technology, patients with LVADs seem to be spending a greater proportion of their days engaging in health care. These quantitative findings critically inform shared decision-making about LVAD therapy, highlighting the need to educate patients and their caregivers about expectations for self-care and potential burden. The results may also lead to implementation of strategies to enhance efficiency in patient-centered care delivery and possibly reduce time burden.

Treatment burden is an important patient-centered quality measure with increasing relevance in the era of value-based care and patient-reported outcomes.²³ Distinct from disease burden or symptom burden, treatment burden describes another aspect of health care workload and the impact of this workload on functional status, HRQOL, and treatment adherence.^24–26 In the general HF population, the concept of “home-time,” defined as the time a patient spends alive and out of a health care institution, has gained traction as a patient-centered outcome and a meaningful clinical trial endpoint.²⁷ The concept of time burden of treatment is complementary to “home-time” and is perhaps more relevant for LVAD patients, for whom time spent in ambulatory care may pose a bigger burden than inpatient time. Thus, although LVAD patients experience improvements in HF symptoms and HRQOL following implantation, this often occurs with the burden of significant time spent in LVAD-related health care activities outside of the home.⁴

Time burden of treatment has been examined in more detail in oncologic care, in which advanced patient age, intensive treatment regimens, and variable prognosis necessitate discussions about treatment tradeoffs and efforts to minimize treatment burden.^25,28 In one study of patients with metastatic pancreatic cancer, patients spent 10% of days survived engaged in ambulatory health care encounters.²⁸ Patients with LVADs face similar challenges to patients with advanced malignancies, including advanced age, multiple comorbidities, risk of adverse events, and variable survival.^3,29–33 Further, we found that patients with LVADs spend a greater proportion of their days engaged in ambulatory health care encounters (15%) than patients with metastatic cancer, highlighting the immense burden of treatment in LVAD therapy.

The reasons for the large amount of time spent engaging in health care among LVAD patients are likely multifactorial. LVAD is an intense therapeutic intervention for a sick patient population that is at least perceived to require intensive surveillance for monitoring of certain aspects of care including anticoagulation, renal function, electrolytes, markers of pump thrombosis, and hemodynamic response to therapy. In fact, “intensive surveillance protocols,” involving a schedule of clinic visits and protocol of routine diagnostics, have been proposed as strategies to achieve improved patient outcomes.³⁴ Patients with LVADs also may suffer adverse events that affect multiple organ systems, often necessitating outpatient consultation with multiple specialists.³ Finally, although significant progress and policy recommendations have been made to coordinate care in diseases such as cancer, there are not yet standardized widely adopted practice standards for care coordination in LVAD care.^35,36

Our study revealed several findings that can be used to inform care. First, the majority of days with health care encounters were days spent in ambulatory care. Thus, although the literature on health care use among LVAD patients has largely focused on emergency department use and hospitalizations, our study suggests that the majority of days spent engaging in health care are spent in the ambulatory setting.^5–8 This finding provides a feasible target for initiatives to reduce burdensome health care visits such as consolidation of clinic visits and tests, enhanced use of telemedicine, and reduced frequency of blood draws if medically appropriate. Second, among patients who survived at least 1 year after LVAD, the proportion of days alive spent engaging in health care was lower after the first year for the total cohort and DT patients. Further, the proportion of days spent engaged in ambulatory encounters was significantly lower after the first year. Although this finding suggests that the time burden of treatment may decrease for patients who survive for longer with LVAD support, the time burden of therapy remains substantial even after 1 year. Finally, our longitudinal analysis revealed that patients with relatively later implantation years may be spending a greater proportion of their days engaged in health care despite improved pump technology. The reasons for this are likely multifactorial, but we suspect are primarily related to the implantation of LVADs in an increasingly older population with more comorbidities, potentially rendering them vulnerable to both more LVAD-related adverse events, including gastrointestinal bleeding, and age-related health issues.^37,38 This hypothesis must be further evaluated in subsequent analyses. It is unclear whether changes in practice patterns, including more frequent and standardized testing and clinic visits, have contributed to these findings. Thus, in addition to efforts to improve pump technology and tolerability over time, initiatives to improve treatment burden are clearly needed.

Study limitations

This study has several limitations. This is a retrospective, single-institution study with a limited sample size. Although data maintained by our center shows that our outcomes (1-year post-implantation survival, readmissions) are comparable to national outcomes based on INTERMACS (Interagency Registry for Mechanically Assisted Circulatory Support) reports, prospective multicenter studies are needed to more accurately quantify the patient experience and account for heterogeneity in practice patterns.³⁹ Additionally, the number of patients with the contemporary Heartmate III device (n = 50, 13%), which is associated with fewer complications, was relatively small. Because of unreliable timestamps in the medical record for most outpatient encounters, we were limited in our ability to capture the actual time spent in ambulatory care. To overcome this, we used national survey data and expert consensus to estimate care time.^28,40 However, it is possible that our assessment of time burden was over- or underestimated. Our adjudication also found missing and extraneous encounters in our dataset when compared to manual review of the EHR. However, given that the majority of these encounters were related to ambulatory bloodwork or imaging procedures, we considered the degree of missingness acceptable and unlikely to significantly impact the overall findings. Finally, our analysis did not capture encounters or days spent in facilities outside the home, such as long-term acute care or inpatient rehabilitation facilities. We also did not capture time spent commuting to health care centers for care. Although it is unclear how time spent in these facilities or commuting affects the patient treatment experience or HRQOL, it is likely tour analysis underestimated the time burden of health care outside the home and could be explored in future studies.

CONCLUSIONS

Our novel study describes times spent engaging in health care for LVAD patients and is a critical step towards an enhanced understanding of the treatment burden in these patients. Ultimately, these data may inform the shared decision-making process for patients and clinicians considering LVAD therapy, as well as initiatives to streamline patient care and minimize burden, including use of telemedicine, consolidation of visits, enhanced care coordination between specialties, and reduction in tests and procedures. Prospective studies that more accurately capture the patient experience, including commuting time, time spent in medical facilities outside the home, and time spent on health care activities inside the home, are needed. Additionally, qualitative work and survey studies examining patient HRQOL as a function of time and treatment burden are essential.

PERSPECTIVES.

COMPETENCY IN MEDICAL KNOWLEDGE:

LVADs were associated with significant time spent engaging in medical care outside of the home after implantation. Patients spent a median 23.2% of their days alive with an LVAD engaging in health care, and a median 15.0% of days engaging in ambulatory care. For patients who survived >1 year with an LVAD, proportion of days spent engaging in health care decreased after 1 year for the total cohort and patients implanted as DT. For all groups (the total cohort, BTT, and DT), a smaller proportion of days were spent engaging in ambulatory health care after the first year than during the first year. However, time spent engaging in health care remained significant, and there was a strong association between year of implantation and proportion of days spent engaged in health care.

TRANSLATIONAL OUTLOOK:

Although LVADs are associated with improved HRQOL and survival, the time patients spend engaging in the health care system after implantation remains significant. These data suggest that ambulatory care specifically represents a viable target for optimizing care efficiency and delivery for patients with LVADs. The data also have implications for patient-centered counseling, whereby certain patients may expect to spend less time engaged in health care after 1 year of survival with an LVAD.

Abbreviations and Acronyms

BTT

bridge-to-transplant

DT

destination therapy

HRQOL

health-related quality of life

EHR

electronic health record

LVAD

left ventricular assist device