From 2016 through 2021, a total of 226 primary LVAD implantations took place, either as bridge-to-transplant, bridge-to-decision or destination therapy. The annual number of implantations increased from 32 in 2016 to 45 in 2020 (p = 0.068) (Fig. 1a). The total LOS around implantation (including clinical screening/optimisation phase and post-operative course) was relatively constant over the years, with a median LOS of 41 days (95% CI: 37–43; p = 0.69) (Fig. 1b). Prior to LVAD implantation, the median LOS was 11 days (95% CI: 10–13), whereas the median LOS after LVAD implantation (intensive care unit and ward combined) was 27 days (95% CI: 25–29) (Fig. 1c). No differences were seen in pre- and post-LVAD implantation LOS over the years (data not shown).

Healthcare consumption in patients undergoing left ventricular assist device (LVAD) in period 2016–2021. a Number of primary LVAD implantations. b Median length of hospital stay (pre- and post-LVAD implantation hospital stays combined). c Overview of LVAD implantation process. d Number of LVAD patients in outpatient clinic (bold line: β: 28; p < 0.001) and total number of outpatient clinic visits (dashed line: β: 114; p = 0.003). e Number of emergency department visits (β: 24; p = 0.003). f Number of readmissions (β: 31; p = 0.001)

Due to the increasing number of implantations in the cohort, the number of patients followed at the LVAD outpatient clinic increased steadily over the years. In 2016, a total of 26 individual patients who had received an LVAD since 1 January 2016 were followed compared with 163 patients in 2021 (β: +28 patients/year; p < 0.001) (Fig. 1d). This resulted in an increase in the number of outpatient clinical visits (defined as a visit to the cardiologist or specialised LVAD nurse) from 124 in 2016 to 812 in 2021 (β: +114 visits/year; p = 0.003) (Fig. 1d), ranging from 1 to 56 visits per patient, with a median of 13 visits during the study period. The annual number of outpatient clinic visits decreased by 1 per patient-year follow-up (B: −0.97; p = 0.002) (Tab. 1).

The number of post-LVAD emergency department visits increased over time (β: +24 visits/year; p = 0.003) (Fig. 1e). In the 6‑year period, 161 individual LVAD patients visited the emergency department, resulting in a total of 553 emergency department visits (median: 2 visits/patient; range: 1–20). Nine patients visited the emergency department ≥ 10 times during the study period. Over time, no differences were found in the number of emergency department visits per patient-year follow up (B: 0.06; p = 0.96) (Tab. 1). Of the emergency department visits. 352 (64%) resulted in hospital admission. Patients presented primarily to the cardiologist or cardiothoracic surgeon (n = 429; 78% of presentations), followed by the neurologist (n = 40; 7%), gastroenterologist (n = 20; 4%) and the ear-nose-throat specialist (n = 20; 4%).

A total of 614 readmissions were recorded in the period 2016–2021. The annual number of readmissions of patients increased from 12 in 2016 to 180 in 2021 (β: 31; p = 0.001) (Fig. 1f). Over time, no differences were found in the number of admissions per patient-year follow up (B: −0.04; p = 0.94) (Tab. 1). A total of 154 individual patients with an LVAD were readmitted after implantation, ranging from 1 to 42 admissions per patient. The median time until the first emergency department visit was 147 days (95% CI: 122–209) after implant, whereas the median time until the first readmission was 167 days (95% CI: 117–209). Table S1 in the Electronic Supplementary Material illustrates the reasons for hospital admission.

Patient characteristics are summarised in Tab. 2. Patients who received an LVAD in the period 2016–2021 were on average 53 ± 13 years of age; this mean age remained constant over the years (data not shown). Of the patients, 66% were male. The underlying aetiology of HF was ischaemic in 67 patients (30%). Most patients received an LVAD while being supported by inotropes. The Interagency Registry for Mechanically Assisted Circulatory Support (INTERMACS) class was 2 in 67 patients (30%) and 3 in 54 patients (24%). Temporary mechanical circulatory support was indicated in 36 patients (16%) in whom the LVAD was implanted. Over the years, no significant differences were seen in INTERMACS classification at LVAD implantation (p = 0.51) (Fig. 2a). There were no gender differences in INTERMACS class (p = 0.807): 12 female patients (16%) were implanted while on temporary mechanical circulatory support compared with 24 of the male patients (16%) (see Table S2 in Electronic Supplementary Material).

a Interagency Registry for Mechanically Assisted Circulatory Support (INTERMACS) classification at left ventricular assist device (LVAD) implantation. b Mean age at LVAD implantation. c Kaplan-Meier curve illustrating survival after LVAD implantation (patients censored after heart transplantation)

LVADs were mainly implanted in patients aged 51–60 years (n = 71; 31%) and those aged 61–70 years (n = 73; 32%). The youngest patient was implanted at age 16 years and the eldest at 74 years of age. The age distribution at implantation remained stable over the years (Fig. 2b), and the mean ± SD age did not differ between men (52 ± 14 years) and women (54 ± 12 years; p = 0.10). There was no significant difference in INTERMACS classification across different age categories (p = 0.31; see Table S3 in Electronic Supplementary Material). Of the patients aged ≤ 50 years, 30% were implanted while on temporary mechanical circulatory support or they had INTERMACS class 1, compared with 17% of those aged > 50 years.

The patients included in our cohort were referred to the UMC Utrecht for advanced HF care from all over the Netherlands. The 6 main referring regions were the provinces North Holland (n = 57; 25%), Utrecht (n = 45; 20%), Gelderland (n = 42; 19%), North Brabant (n = 32; 14%), Overijssel (n = 19; 8%) and South Holland (n = 13; 6%).

Figure 2c displays a Kaplan-Meier curve illustrating survival after LVAD implantation, with patients being censored in the event of heart transplantation.

The median duration of hospital admission around the LVAD implantation was 41 days. Hospital stay was slightly longer with increasing age (β: +0.26 days/year; p = 0.01) (Fig. 3a). A similar trend was seen for duration of pre-LVAD and post-LVAD hospital stay (data not shown). A trend towards a longer hospital stay was observed for women compared with men, although this difference was not statistically significant (median: 46 days; 95% CI: 42–50 for women vs 38 days; 95% CI: 36–42 for men; p = 0.17). This was mainly determined by a longer post-LVAD implantation LOS in women than men (30 days; 95% CI: 29–36 vs 24 days; 95% CI: 23–27; p = 0.053). The pre-LVAD implantation duration was similar for men and women (11 days; 95% CI: 10–14 vs 11 days; 95% CI: 9–14; p = 0.68) (Fig. 3b).

a Scatterplot of length of hospital stay around left ventricular assist device (LVAD) implantation (pre- and post-implantation) in relation to age (β: 0.26; R2: 0.03; p = 0.01). b Median length of hospital stay (with 95% confidence interval (CI)) pre- and post-LVAD implantation in men and women. c Median length of hospital stay (with 95% CI) in pre-LVAD implantation phase (p = 0.001) and post-LVAD implantation phase (p = 0.001) in relation to Interagency Registry for Mechanically Assisted Circulatory Support (INTERMACS) class. T temporary mechanical circulatory support

The relation between the INTERMACS classification and duration of hospital admission is summarised in Fig. 3c and Table S4 in the Electronic Supplementary Material. Patients who received an LVAD while on temporary mechanical circulatory support (INTERMACS class T) stayed significantly longer in hospital (median: 60 days; 95% CI: 48–65) than patients with INTERMACS class 1, 2, 3 or 4/5 (median: 36, 36, 39 and 44 days, respectively; all p < 0.05). Pre-LVAD LOS differed by INTERMACS class (p < 0.001). The median pre-LVAD hospital stay was 5 days (95% CI: 1–16) in patients with INTERMACS class and 16 days (95% CI: 12–20) in those with INTERMACS class 4 (p = 0.15). There was an inverse relation between post-LVAD implantation hospital stay and INTERMACS class (p = 0.001), which mean that the lower the INTERMACS class, the longer the post-LVAD hospital stay. Patients on mechanical supported stayed 43 days (95% CI: 35–54) in hospital compared with 23 days for those with INTERMACS class 4 (95% CI: 21–28; p = 0.001).