Before engaging potential study participants, the intervention evaluation protocol was approved by the Research Ethics Board of the University Health Network (Reference No: 16-5136). Ethics approval was secured from 77 sites, including hospitals, Community Care Access Centers, community health centers, and family health teams participating in the telehomecare program across 8 local health integration networks (LHINs). The evaluation study team contacted potential patient participants by telephone, and if they agreed to participate, verbal consent was obtained using a telephone consent script approved by the Research Ethics Board. All other stakeholders who agreed to take part in the study provided consent either in person or by telephone. Participants who provided consent received a copy of the study information and consent form and were informed that the research findings would be published in reports, articles, and presentations. All study data were anonymized and deidentified.

This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines [27]. This study used a prospective longitudinal cohort and cross-sectional design, and the population included patients with HF or COPD enrolled in the telehomecare program, telehomecare nurses, and informal caregivers participating in telehomecare across 8 LHINs in Ontario from July 2016 to December 2019. The LHINs were created to plan, coordinate, integrate, and fund health services at the local level and are separated into different geographical catchment areas. Patients and caregivers were referred to the program through the LHINs or self-referrals and were screened for program eligibility criteria by telehomecare staff. Patients and caregivers were approached and purposively selected to ensure variation in sampling from each LHIN (urban vs rural) and disease type. Those patients and caregivers who agreed to participate were administered longitudinal surveys. The OTN, in collaboration with the LHIN engagement leads, referred telehomecare nurses to the evaluation study team. Nurses who were delivering telehomecare at the time of the evaluation were considered and approached to participate in the cross-sectional surveys.

Patients and caregivers qualified with a documented HF or COPD diagnosis, with or without comorbidities. “Heavy users” had recent respiratory or cardiac hospitalizations or multiple emergency visits. They also may have been receiving nursing services or had frequent primary care visits. Eligibility included being aged 18 years and older, capable of consent, fluent in English, able to operate telehomecare equipment, and residing in a private or retirement home with a landline. Additional criteria for the intervention study required consent to telehomecare, sharing contact and health information, and participating in evaluations. Patients were contacted and enrolled in surveys within 2-3 weeks of their telehomecare start date. Health care providers were included if they had referred a patient who had previously been enrolled in the telehomecare and/or had participated in delivering telehomecare as a provider or care administrator. For surveys, nurses needed at least 2 months of experience in providing care using telehomecare.

Patients aged 18 years or younger, those without an established diagnosis of COPD or HF, those unable or unwilling to provide verbal informed consent, and those demonstrating nonadherence to the telehomecare program were not eligible. The telehomecare nurse worked with each patient on a case-by-case basis to assess their willingness to participate in their own care and to determine if missed consultations and their reasons reflected overall nonadherence. Additionally, individuals who could not or would not use telehomecare equipment, or who did not have a regular caregiver to assist with the equipment if needed, were also excluded. Health care providers who did not practice in any of the 8 LHINs were excluded.

The telehomecare program intervention has been described extensively elsewhere [22]. Telehomecare aims to improve self-management skills for patients with HF and those with COPD by providing remote monitoring and nurse-delivered individualized health coaching sessions using the 5A’s (assess, advise, agree, assist, and arrange) Behavior Change Model [23,28,29]. Patients are provided with a touch-screen tablet, weight scale, blood pressure monitor, and pulse oximeter for weekday submission (Monday-Friday) of clinical information and self-report [23,30]. Telehomecare nurses monitor incoming patient information and respond to any deteriorating conditions or alerts triggered due to data received outside “normal” thresholds [31]. Following an alert, a nurse follows up and determines a course of action. If an alert impacts a patient’s well-being, telehomecare nurses immediately communicate with the patient’s care team.

The outcomes for the longitudinal surveys were patient-perceived HRQoL, disease-management skills, perception of telehomecare, satisfaction with telehomecare, and caregiver-perceived strain. Cross-sectional surveys were conducted to assess nurse perception and satisfaction with telehomecare. Patients and caregivers were administered telephone surveys at baseline, 3, 6, and 12 months after starting the program from July 2016 to December 2019. Participants were contacted for a follow-up at least 4 (maximum 10) times and within 1 week before and 1-2 weeks after the scheduled date. Validated questionnaires were used to evaluate patients’ general and disease-specific HRQoL, satisfaction with telehomecare, self-management skills, and telehomecare perception. A complete list of survey instruments and details can be found in Multimedia Appendix 1.

Missing values at the survey level were imputed for the MLHF using its predetermined methods to neutralize the effects of missing values when calculating and comparing changes in scores [41]. Item number 10 within the MLHF was found to have a large proportion of item-level missingness, likely due to the item reflecting sexual activities. The TMPQ can only be scored when all items have been completed. Responses missing completely at random for the TMPQ were imputed using the last recorded observation for that question carried forward [42]. Missing values within the SOLD, SES, CSQ-4, MCSI, and NSQ were not imputed within the dataset, as this did not impact the ability to score the surveys. Missing values in the SF-12 were not imputed, and incomplete responses were not analyzed.

All analyses were performed using SAS Statistical Software (version 9.4, SAS Institute Inc). Descriptive statistics were used to summarize patient, provider, and caregiver survey data. Continuous variables were described using means and SDs. Categorical variables were described using counts and percentages. The entire cohort of patients was used to analyze the survey data for the SF-12, EQ-5D, SES, TMPQ, and CSQ-4. To look at telehomecare’s impact on the varying disease types (HF vs COPD), analyses using the MLHF and SOLD were performed using the respective patient cohorts. Repeated measures of the patient and caregiver surveys for continuous values were analyzed using generalized linear mixed models to determine if outcome measures of HRQoL, client satisfaction, telemedicine perception, and self-management scores were changing over time. A separate analysis was conducted for each outcome measure. The models were selected using the Akaike Information Criterion and used Restricted Maximum Likelihood estimation methods. Covariates included LHIN, disease type, age, gender, time, and the interactions between these variables of interest (time × diagnosis and time × gender). The Kenward-Roger approximation was used for the degrees of freedom, and the α level was set at 0.05 for all parameter tests [43]. EQ-5D-3L utility index scores were calculated using the Canadian value set [44]. Additionally, to address potential response bias, patient demographics of participants who consented and those who declined participation were analyzed using chi-square and 2-tailed Welch t tests.

Between July 2016 and December 2019, a total of 318 participants consented to take part in the study. Of those, 38 participants were excluded because they subsequently declined the surveys, withdrew, and were nonresponsive when contacted. In total, 194 patients (HF, n=117; COPD, n=77), 62 caregivers, and 24 nurses were enrolled in the program. An overall participant response rate of 51% (280/551) was observed. Figure 1 outlines the flow of participants through the study. The average age of patients with HF and those with COPD who participated in the surveys was 71 (SD 11.3) years and 70 (SD 11.1) years, respectively, and 52% (100/194) were men.

To assess potential response bias, we examined the characteristics of patients who consented to surveys and those who declined. The declined group consisted of patients who refused participation, whether they were approached directly or identified through the Patient Management and Monitoring System. There were no differences seen in relation to gender and condition; however, patients who declined to participate in surveys were older adults (Table 1). There is a possibility that older patients may have been in worse health and may not have benefited from being enrolled in telehomecare.

We administered longitudinal survey instruments at baseline, month 3, month 6, and month 12 for patients and caregivers; however, not all participants completed the surveys at all timepoints. Cross-sectional surveys were administered to nurses with at least 2 months of experience in providing care using telehomecare. As noted in Table 2, the administration of surveys declined over the study duration due to participants being nonresponsive or withdrawing at various timepoints; a patient could complete the baseline surveys but fail to complete subsequent follow-ups. Across each instrument 40% (47/117; MLHF) to 42% (47/112; SES, TMPQ, and CSQ) of patients with HF completed surveys at all timepoints, 34% (26/77; SOLD) to 36% (27/76; SES, TMPQ, and CSQ) of patients with COPD completed surveys at all timepoints, and 37% (23/62; MCSI) of informal caregivers completed surveys at all timepoints.

In this study, participation in telehomecare was shown to significantly improve perceived HRQoL of patients with HF across all timepoints using the MLHF. An MCID was observed in each domain of the MLHF, inferring a clinical improvement in HRQoL of patients with HF. Additionally, MCIDs were observed in the SOLD, indicating that patients with COPD experienced clinical improvements in physical, coping, and treatment satisfaction function. Results from the EQ-5D indicated that patients with COPD had significantly higher VAS scores than patients with HF. Furthermore, MCIDs were detected in the SF-12 HF PCS scores, indicating a clinical improvement in the health of patients with HF. Using the SES instrument, patients reported being confident in self-managing their diseases throughout the entire study duration, and satisfaction with telehomecare was consistently positive using both the TMPQ and CSQ. Patients who were involved in telehomecare perceived it to be important, but once they completed the program and upon follow-up, their perception of telehomecare was understandably lower. While satisfaction and perception of telehomecare were highly rated at months 3 and 6, as time went on, lower scores were reported by older patients. This could be due to the older patients in the program being in poorer health and having a more difficult time self-managing their comorbid conditions.

There are only a few studies which provide patient-reported outcomes as a primary measure or as well-explored secondary outcomes within the literature. The literature surrounding the use of telehealth interventions and their ability to impact patients’ HRQoL is inconsistent in definitively proving its value in patients with HF and those with COPD. A review examining telemonitoring in patients with HF and those with COPD further identified conflicting results on telehomecare’s impact on HRQoL [14].

Conclusions from the Telemedicine for Heart study identified in the review indicated that patients with HF who received telemonitoring improved their quality of life as measured by the Short-Form-36 (SF-36) [14,45]. Conversely, upon further review of the study, both groups’ scores were not significantly different from one another, SF-36 data were not reported completely, and some telemedicine group scores were worse than controls. Therefore, the conclusion that HRQoL was improved in patients with HF could not be supported in the Telemedicine for Heart study [45]. The Telemedicine for Heart study population (mean age 65.2, SD 9.9 years; 233/281, 83% men) varied from this telehomecare study as participants were, on average, younger and their study population was less diverse with only 48 (17%) women [45]. The BEAT-HF 180-day MLHF score (28.50), closely reflects the month 6 score observed in this telehomecare study (29.1, SD 21.4). Additionally, the BEAT-HF telehomecare intervention and patient population were similar to this study with a median age of 73 (range 62-84) years and 46.6% (664/1437) female participants. However, the BEAT-HF trial did not provide HRQoL data in full (no baseline data) or report any measures of dispersion, limiting the generalizability of this study’s findings [46]. The Medly telemonitoring intervention reported improvements in MLHF scores of patients with HF in comparison to controls over 6 months. The telemonitoring group was predominately men (78%) and was significantly younger (mean age 58, SD 15.5 years) than the telehomecare patient group. Additionally, based on the baseline MLHF overall domain scores of the telemonitoring group (53.2, SD 26.3), these patients had worse symptoms than those enrolled in telehomecare (44.6, SD 23.8) [47].

A Cochrane review of 11 telemonitoring studies that measured HRQoL in patients with HF stated that only 5 studies reported statistically significant improvements in HRQoL outcomes [48]. Among the studies that reported improvements in HRQoL of patients with HF, there are a few caveats to take into consideration. Antonicelli et al [49] reported improvements only in the SF-36 health perception subscale, but component scores remained nonsignificant. The Medicare Coordinated Care Demonstration (MCCD) Project improved SF-36 PCS, MCS, and MLHF overall, physical, and emotional domains of patients with HF. However, these improvements were not significantly different from the results seen in the control group [50]. MCCD patients received a similar remote monitoring program as those in telehomecare. However, their patients with HF were older (mean age 73, SD 8 years) than those enrolled in telehomecare and were predominately male (71/101, 70%). The MCCD MLHF domain scores at baseline, month 6, and after 1 year were also compared to MLHF scores observed in this study [50]. The TIM-HF study reported improvements in SF-36 physical functioning scores in the intervention group versus the control group at 12- and 24-month follow-up after receiving remote patient monitoring. The TIM-HF study patient population was younger than telehomecare, with a mean age of 66.9 (SD 10.8) years, and was less diverse, with primarily men (81%) enrolled [51]. Subsequently, the TIM-HF II study, with a mean age of 70 (SD 11) years, like telehomecare, reported no statistical differences in MLHF scores between remote patient monitoring and usual care from baseline to 12 months [52].

Two Cochrane reviews examining patients with COPD with a history of admissions for exacerbations concluded that telemonitoring provided inconclusive evidence of a benefit and did not improve patients’ quality of life compared to usual care [53,54]. The review of tele–health care interventions for COPD by McLean et al [55] identified only 2 studies which reported HRQoL using validated instruments. The meta-analysis in this review reported an imprecise pooled estimate, which indicated insufficient evidence to demonstrate a clear benefit of tele–health care interventions for patients with COPD. The included studies contained patients aged 69.4 (SD 6.5) years and those aged 70 (SD 9) years, which closely related to the telehomecare population [56,57]. The TeleCare North study was a large-scale trial aimed at establishing evidence of HRQoL in patients with COPD receiving tele–health care in Denmark and closely resembling telehomecare. TeleCare North reported nonsignificant differences between intervention and control groups when looking at SF-12 PCS and MCS scores over 12 months. Interestingly, the findings of the TeleCare North study suggest that tele–health care interventions may provide a psychological benefit and work to slow the deterioration of mental health in patients with COPD [58]. In comparing the unadjusted MCS scores from baseline to month 12 for patients with COPD, a similar reduction in change scores was observed (telehomecare, –1.7 vs TeleCare North, –1.5).

These results reinforce the inconsistencies in the literature regarding telemonitoring improving HRQoL among patients with HF and those with COPD [58,59]. Using a disease-specific measure such as the MLHF may be more appropriate than the SF-12 or SF-36 in capturing significant changes in the health of patients with HF over time, while a more responsive disease-specific measure may need to be identified for longitudinal use in patients with COPD. Additionally, a greater emphasis may need to be placed on the MCID of HRQoL for patients with HF and those with COPD receiving telehomecare. Telehomecare seems to be well received by the majority of patients, but home telemonitoring may have more significant benefits to the HRQoL of patients with HF and those with COPD in mild to moderate disease states than those whose health has already begun to deteriorate [60]. These results provide more credence to the use of a risk stratification framework to identify patients with mild to moderate risk of hospitalization who should be directed to telehomecare [61,62]. Future research should focus on using disease-specific measures to assess HRQoL in patients and assessing change using validated MCIDs to identify meaningful change.

For telehomecare to be readily accepted and implemented, health care providers must be on board and believe it brings tangible benefits to their patients. Engagement and collaboration from clinicians may be critical success factors for telehomecare, as clinicians are at the forefront of the care provided to patients, and they should be a part of designing the program. Additionally, telehealth interventions should be integrated into the clinician’s workflow, creating an environment that is conducive to the use of the equipment [62-64]. El-Dassouki et al [65] report how a telemedicine intervention can reduce uncertainty, reduce caregiver strain by easing their workload, and mitigate clinical knowledge requirements.

A vital strength of the longitudinal surveys was the use of validated generic and disease-specific measures, which provided us with the opportunity to compare important changes in patient’s health over time. Using the MLHF and SOLD questionnaires allowed us to measure fundamental changes that patients deemed as important to their health status regarding their condition. The length of follow-up (12 months) was also an asset for the study because we were able to evaluate patients’ self-management skills and perception of telehomecare after being off the program for a considerable amount of time. This study has limitations, which warrant discussion. First, this study was a part of a comprehensive program evaluation, which did not have a control group to evaluate the impact of telehomecare compared to usual care. Second, we were unable to gain access to a proportion of the patient’s demographic information due to the migration of OTN telehomecare databases during this study’ period. Third, there was no longitudinal follow-up for the NSQ; scores were cross-sectional. Fourth, we were not notified each time a patient passed away within the program; consequently, not every patient death was explicitly documented. When a participant was found to have died, their status was recorded as a withdrawal. Lastly, we were unable to recruit a reasonable number of physicians to take part in the study and gain their perceptions of telehomecare, which is critically important in understanding its adoption by health care providers.

Telehomecare was created to improve access to quality care and reduce health care use by delivering health care at a distance and providing tailored feedback. The OTN telehomecare program has demonstrated an ability to improve the HRQoL of patients with HF and those with COPD. However, the long-term sustainability of HRQoL improvements requires further investigation. Additionally, telehomecare has been shown to decrease informal caregiver perceived strain, and telehomecare nurses described moderate levels of satisfaction and perceived quality of care with telehomecare. Telehomecare seems to be well received by most patients with HF and those with COPD, but more research is needed into home telemonitoring to determine its value in improving HRQoL.