There is strong evidence that rehabilitation programs using home telehealth can be as or more effective than usual care. These programs almost invariably involve regular real-time interaction between patient and provider. Feedback is immediate rather than asynchronous. In this respect, rehabilitation programs using home telehealth bear a closer resemblance to their conventional counterparts than remotely delivered oral anticoagulation therapy and emotional support interventions.
There is moderate evidence that home telehealth can be used to good effect in the delivery of oral anticoagulation therapy. Several studies show increased time in therapeutic range and reduced levels of variability in international normalized ratio (INR) with use of remote monitoring programs. The evidence suggests that high engagement and corresponding improvement in self-care activities may be essential to realizing these benefits. Odds of success appear higher when the intervention design includes frequent patient reminders and other components designed to improve adherence to protocol.
A recent study of hypertensive stroke patients found that intensive monitoring through a home telehealth unit was also effective in increasing compliance with daily blood pressure monitoring and decreasing the percentage of patients with abnormal blood pressure. As this study was retrieved in a final scan of 2011-2012 literature and was not subject to the same level of analysis as the other studies included in this review, we will not attempt to extrapolate beyond the conclusions provided by its authors. See Kuo et al. (2012) for more details.
At present, there is no evidence that home telehealth can improve emotional well-being. This lack of evidence is primarily attributable to a shortage of research rather than a large body of negative findings. However, the few studies measuring depression, self-perceived health status, and quality of life found that home telehealth had no effect on patients’ scores.
Evidence on home telehealth’s effects on mortality and complication rates is insufficient. Only one study attempted to measure these outcomes. While no intervention effect was found, the fact that some instantiations of home telehealth significantly improve clinical outcomes suggests that further research may result in different conclusions.
In general, patient uptake of home telehealth was high. Few had trouble with the technologies used in the interventions. These included point-of-care blood testing meters, internet-enabled computers, and mobile devices for transmitting information to providers. Satisfaction levels, when reported, were in the 80-100% range. However, these findings should be interpreted with caution. Those who consent to take part in a home telehealth program may be more comfortable with technology than most. It is possible that the populations featured in these studies are not representative of the typical stroke patient.
Little information on patient costs and time savings was available.
Uptake and Use of Technology
Summary: Uptake and use of technology were reported in 4 studies (Lutz et al., 2009; Pierce et al., 2009; Ryan et al., 2009; Salvador et al., 2008). In general, participants appeared to be satisfied with the usability of the intervention technologies. However, it remains challenging to draw conclusions about the acceptability of technology to stroke patients. Two-thirds of the studies retrieved did not report on uptake of technology or conflated uptake with completion of the intervention. In addition, there were significant differences in the technologies used in studies that did examine uptake.
Findings also need to be interpreted with an eye to factors such as participants’ computer literacy, which was reported only in 1 study (Pierce et al. 2009), as well as access to technology, which was a criterion for inclusion in 2 studies (O’Shea et al., 2008; Ryan et al., 2009). The populations included in these studies may not be representative of typical stroke patients. Patient uptake and use of technology cannot reasonably be considered independent of access and technological literacy, and eligibility criteria that include access to technology may produce sampling bias.
Study Details: At the endpoint of Lutz et al. (2009), 87% of patients reported finding the educational component of the telehealth intervention helpful. In Ryan et al. (2009), 99% of patients who completed the satisfaction survey (117 of 132 participants) found the point-of-care INR testing meter easy to use. Salvador et al. (2008) reported that 52 of 54 participants were able to use the requisite coagulometer and cellular phone to self-test and report their INR. It was further reported that 45.5% of INR tests were completed on the scheduled day and 74.8% within 3 days. Pierce et al. (2009) examined the effectiveness of a web-based intervention on the well-being of stroke caregivers and found that, in general, participants in the intervention group reported satisfaction with the “Caring~Web” website, and used it for an average of 1 to 2 hours per week over the 12-month study period
Self-Management, Self-Efficacy, and Behaviour Change
Summary: Self-management, self-efficacy, and behaviour change outcomes were reported in 2 studies: one Level 2 evidence (Ryan et al., 2009) and the other Level 4 evidence (O’Shea et al., 2008). Both measured changes in frequency of INR testing. There is moderate evidence supporting an association between home telehealth and significantly increased rates of INR self-testing.
Study Details: Ryan et al. (2008) found a significant difference in INR testing frequency when comparing intervention and control groups: patients using telehealth measured their INR roughly once every 4.6 days on average, while those in the control group averaged once every 19.6 days (p<.001). O’Shea et al. (2008) found that patients measured their INR approximately 4 times more frequently during the 6-month intervention period than during the 6 months preceding.
Recent Developments: A scan of material from 2011-2012, a time period not covered by our initial searches, found 1 article that addressed self-management, self-efficacy, and behaviour change outcomes (Kuo et al., 2012). Significantly increased rates of daily blood pressure monitoring were seen in stroke patients enrolled in an intensive home monitoring program. This study’s pre-/post-test design was a limitation. As this study was not subject to the same level of analysis as the other studies included in this review, we will not attempt further analysis. See Kuo et al. (2012) for more details.
Clinical Outcomes, Symptoms, and Health Status
All 8 studies reported on clinical outcomes and/or symptoms and health status outcomes. The most commonly used measures in studies on post-stroke rehabilitation were instruments designed to assess upper limb function. These included the Fugl-Meyer scale, the Abilhand scale, and the Ashworth scale. In the studies that focused on educational interventions, the authors report primarily on depression and self-rated health of caregivers. Time in therapeutic range and INR result variability are the most frequently reported outcomes in interventions used to support oral anticoagulation therapy management.
Summary: Post-stroke physiological outcomes were reported in 3 studies: two Level 2 studies (Piron et al., 2006, 2009) and one Level 4 study (Holden et al., 2007). There is strong evidence that significant improvement in physiological functioning can occur when using home telehealth and moderate evidence that improvement on selected measures can be significantly greater than that seen with usual care.
Study Details: A Level 2 study by Piron et al. (2006) reported significant improvements in Fugl-Meyer scores in both intervention and control groups (IG: 48.8 (SD 7.1) to 53.3 (SD 7.2) post-intervention to 52.8 (SD 7.1) at follow-up; CG: 48.2 (SD 4.2) to 50.5 (SD 4.6) post-intervention to 49.9 (SD 5.1) at follow-up). The intervention group also showed significant improvements in Ashworth Scale scores (2.1 (SD 1.1) at baseline to 1.3 (SD 1.1) post-intervention to 1.6 (SD 1.1) at follow-up). No corresponding improvement was seen in the control group. Between-group comparisons showed no significant differences between groups.
In Piron et al. (2009), also Level 2 evidence, the authors report that post-intervention Fugl-Meyer scores and were significantly higher in the intervention group than in the control group (UE scores of 53.1 (IG) vs. 48.8 (CG). Significant improvements in the Abilhand scale and Ashworth scales were reported for both intervention and control groups. However, no significant difference was found between groups. On the Abilhand scale, baseline, post-intervention, and follow-up scores increased from 23.3 (SD 6.9) to 25.0 (SD 7.6) to 25.7 (SD 7.3) and from 27.3 (SD 4.5) to 28.4 (SD 4.1) to 28.2 (SD 4.0), respectively.
Significant improvements in Fugl-Meyer scores were found in the single-group Level 4 study by Holden et al. (2007) after 15 sessions (effect size 7%, p=.003), after 30 sessions (effect size 17%, p < .0001) and at the 4 month follow-up (effect size 20%, p= .001). Significant improvements were also reported for 3 other indicators of upper extremity function: the Wolf Motor test, shoulder flexion strength, and grip strength.
See table below for additional details (Table C.2.3.1: Patient Outcomes – Rehabilitation).
Self-Rated Health and Depression
Summary: Self-rated health and depression were reported in 1 multi-part study by Pierce et al. (2009) and Steiner et al. (2008). This study was considered Level 2 evidence. At present, there is no evidence that home telehealth reduces or prevents depression in stroke patients or their caregivers.
Study Details: No significant differences were found between intervention and control groups in caregiver self-rated health (Steiner et al., 2008) and no significant difference was found between patient groups in levels of depression (Pierce et al., 2009).
Time in Therapeutic Range
Summary: Time in therapeutic range for patients receiving warfarin therapy for the primary or secondary prevention of stroke was reported in 3 studies: two Level 2 evidence (Salvador et al., 2008; Ryan et al., 2009) and one Level 4 evidence (O’Shea et al., 2008). Ryan et al. (2009) and O’Shea et al. (2008) offer moderate evidence of the ability of home telehealth to improve time in therapeutic range. However, the evidence base is not entirely consistent on this point: 1 Level 2 study showed no significant difference between control and intervention groups.
Study Details: Ryan et al. (2009) report significantly greater median time in therapeutic range during the intervention phase of a cross-over study (effect size 15%, p < 0.001). O’Shea et al. (2008) also found that patients spent a significantly greater mean time in therapeutic range during the intervention period (effect size 11.4%, p < .005). Salvador et al. (2008), however, found no significant differences between groups in time in therapeutic range.
Ryan et al. (2009) and O’Shea et al. (2008) appear to have studied the same home telehealth system, or minor variations thereof. Salvador et al. (2008) used a different platform, but it appeared to have very similar functionality. It is possible that difference in uptake contributed to these divergent findings. Patients in Ryan et al. (2009) and O’Shea et al. (2008) performed self-testing significantly more frequently when using the intervention, and received same-day contact from providers if they missed a scheduled data upload. Their exclusion criteria filtered out those with missed clinic appointments, who might be expected to be less compliant. In Salvador et al. (2008), though the authors state that adherence to protocol was high, under half (45.5%) of INR tests were completed on the assigned day. Just under three-quarters (74.8%) were completed within the 3 days before or after that day. Whether this was related to the design of the study, the characteristics of the sample (mean age was significantly higher), or another factor cannot be determined with certainty.
See table below (Table C.2.3.2: Patient Outcomes – Disease Management) for additional details.
Variability in International Normalized Ratio (INR)
Summary: Variability in International Normalized Ratio (INR) was reported in 2 studies, one Level 2 evidence (Ryan et al., 2009) and one Level 4 evidence (O’Shea et al., 2008). There is moderate evidence associating home telehealth interventions with significant improvement in variability of INR test results.
Study Details: Ryan et al. (2009) report significantly fewer extreme INR test results with the telehealth intervention (effect size 4.3%, p < 0.001). O’Shea et al. (2008) found a significant improvement in INR test result variability during the intervention period (SD .49 vs. SD .57, p< .0001). Possible implications of the differences in the design and implementation of these studies can be found in Time in Therapeutic Range, above.
See table above (Table C.2.3.2: Patient Outcomes – Disease Management) for additional details.
Mortality and Hemorrhagic Complications
Summary: Mortality and hemorrhagic complications were measured in one Level 2 study by Salvador et al. (2008). No evidence of intervention-related reductions in mortality or hemorrhagic complications was found.
Study Details: See table above (Table C.2.3.2: Patient Outcomes – Disease Management) for additional details.
Recent Developments: A scan of material from 2011-2012, a time period not covered by our initial searches, found 1 article that addressed clinical outcomes, symptoms, and health status (Kuo et al., 2012). In this study, stroke patients enrolled in an intensive home monitoring program. The percentage of patients with abnormal blood pressure decreased over the course of the intervention. This study’s pre-/post-test design was a limitation. As this study was not subject to the same level of analysis as the other studies included in this review, we will not attempt further analysis. See Kuo et al. (2012) for more details.
Quality of Life
Summary: One Level 2 study reported on quality of life, and this was not among stroke patients but among their caregivers (Pierce et al., 2009). There is no evidence at present of intervention-related improvements in quality of life among stroke caregivers.
Study Details: Pierce et al. (2009), using the Satisfaction with Life Scale, found no significant differences in quality of life between intervention and control groups or within groups over time.
Cost and Time Savings
No cost or time savings outcomes were reported.