Advice
Clinical and technical evidence
Clinical and technical evidence
A literature search was carried out for this briefing in accordance with the interim process and methods statement for medtech innovation briefings. This briefing includes the most relevant or best available published evidence relating to the clinical effectiveness of the technology. Further information about how the evidence for this briefing was selected is available on request by contacting mibs@nice.org.uk.
Published evidence
Seven studies are summarised in this briefing including 2,110 people. The studies are prospective cohort studies and come from 3 full text articles and 4 conference abstracts. There are also development and validation studies for the TriageHF risk score (a component of the TriageHF Plus care pathway) that are not summarised in detail in this briefing (Burri et al. 2018, Cowie et al. 2013, Gula et al. 2014, Koehler et al. 2019 and Whellan et al. 2010).
There are further abstracts evaluating the TriageHF Plus care pathway that are not summarised in this briefing because they are based on analyses of studies already captured below, or are lacking in detail (Ahmed et al. 2020a, Ahmed et al. 2018, Virani et al. 2016a, Virani et al. 2016b, Virani et al. 2016c and Zieroth et al. 2016).
The clinical evidence and its strengths and limitations are summarised in the overall assessment of the evidence.
Overall assessment of the evidence
The evidence base for the technology is of low to moderate methodological quality. Six of the 7 included studies were done in the UK and so are generalisable to the NHS. The evidence mainly comes from observational studies done by the same research group, so the samples may have some overlap between publications. Some evidence is published in poster abstract form so detail is limited.
The evidence suggests that a high heart failure risk score combined with a telephone assessment may be a useful screening pathway for people at risk of heart failure or with worsening heart failure. However, none of the included studies were randomised and only 1 study made a comparison of TriageHF Plus with standard care (via an unselected historical control group). All 7 studies contained a telephone assessment, but some did not need this and screening questions were not standardised. Three of the studies had authors who had received funding from the company or were company employees. The evidence base would benefit from randomised controlled trials comparing the technology with standard care, focused on cardiovascular and all-cause unplanned hospitalisations, death and other long-term health outcomes.
Ahmed et al. (2021)a
Intervention and comparator
TriageHF Plus compared with a historical control group receiving standard care.
Key outcomes
In the TriageHF Plus cohort there were 135 unplanned all-cause hospitalisations, compared with 358 in the standard care cohort. Rates per person per year for all-cause admissions (0.41 versus 0.57, p=0.001), cardiovascular admissions (0.14 versus 0.29, p=0.037) and heart failure admissions (0.02 versus 0.07, p=0.003) were all significantly lower in the TriageHF Plus cohort.
Ahmed et al. (2019)
Intervention and comparator
TriageHF Plus high heart failure risk score (HFRS) with telephone assessment compared with low or medium HFRS with telephone assessment.
Key outcomes
Over 27 months, 118 people had a high HFRS. There were 113 people with a low or medium HFRS. Telephone contact was made in 127 out of 157 (80.9%) high HFRS cases. At telephone consultation, 90 out of 127 contacts (70.9%) had worsening heart failure or another acute medical problem. Out of 127 contacts, 36 had no apparent cause for high HFRS, but the telephone contact prompted interventions in 8 cases. In the low or medium HFRS comparator group, 98 out of 113 people (86.7%) could be contacted. Of these, 1 person reported worsening heart failure symptoms. No other acute medical issues were identified. The telephone questions specified in the study protocol had 98.6% concordance with the healthcare professional's judgement on whether a person had worsening heart failure. In the 71 people who had either isolated worsening heart failure (n=64), or worsening heart failure with a concurrent medical problem (n=7), the sensitivity of a high HFRS to identify worsening heart failure was 98.6% (92.5 to 100.0%) and specificity was 63.4% (55.2 to 71.0%). Overall accuracy was 74.7% (68.5 to 80.2%).
Strengths and limitations
This is a real-world evaluation of the TriageHF Plus care pathway. The study did not collect outcome data beyond the time of assessment and cannot be used to understand the impact of TriageHF Plus on outcomes such as unplanned hospitalisation and death. It also cannot compare outcomes for people with high HFRS alerts who were contacted versus those who were uncontactable.
Garner et al. (2022)
Study size, design and location
Prospective single centre cohort study of 188 people with Medtronic CIED in the UK.
Key outcomes
Over 24 months, 367 high HFRS alerts were received for 188 people. The mean number of alerts per person was 1.95, and 44 (23%) people had more than 3 high alerts. Contact was made for 303 (87%) alerts. No intervention was needed for 128 (35%) alerts (68 were asymptomatic, 49 were improving clinically and 11 had previously been actioned). There were 53 (28%) unplanned hospital admissions within 6 weeks of the high alert, and 24 of these were for heart failure decompensation. A total of 33 (18%) people died during follow up. The authors concluded that people with high-risk alerts are comorbid and have significant healthcare resource use. The study described the effect on workload, noting that there was only a small number of high-risk alerts when these were shared between different teams. Some users felt that they became unfamiliar with the care pathway, which led to reduced staff engagement.
Strengths and limitations
This study was done in the UK and is generalisable to the NHS. It gives an insight into the implementation of TriageHF Plus within the NHS after 2 years, including the impact on workload. The main limitation is the non-comparative design. Without a control, the impact on outcomes cannot be attributed to TriageHF Plus. One author has received fees from Medtronic for educational activities, grants and consultancy work. Two authors have received honoraria from Medtronic for training and educational activity.
Virani et al. (2018)
Key outcomes
After 8 months, 87 people were included in the study, 5 people withdrew because of system modifications, 3 people died and 5 people withdrew for other reasons or were lost to follow up. During the study, 24 high HFRS alerts were received, resulting in telephone contact in 16 people. Symptoms and behaviours associated with worsening heart failure were seen in 83% of telephone interviews. Excluding non-compliance, symptoms and behaviours of worsening heart failure were identified in 63% of telephone interviews. During the study, 31 medium risk alerts (8.4%) resulted in a telephone assessment based on healthcare professional's discretion. Excluding non‑compliance, 25 interviews showed 1 symptom or behaviour of worsening heart failure. At the beginning and end of the study, HFRS‑estimated risk burden correlated well with healthcare professional's clinical assessment of heart failure risk in the next 90 days (p<0.05 for baseline and exit visit). Interventions occurred in 13 (54%) people with high‑risk scores and 24 (6.4% of all and 77% of contacted) people with medium risk scores. A total of 28 hospitalisations happened during the study, 13 of which were thought to be heart-failure related.
Strengths and limitations
This study suggests that a high HFRS correlates with signs, symptoms and behaviours associated with worsening heart failure. The study did not need a standardised telephone assessment in response to a high-risk score, or a minimum alert frequency, so clinical status may have differed between alerts and it does not adhere to TriageHF Plus requirements. Two authors received research grants from Medtronic and 3 authors are employees of Medtronic.
Ahmed et al. (2021)b
Study size, design and location
Prospective real-world evaluation of 415 people with Medtronic CIED in the UK.
Key outcomes
Over 15 months, 159 TriageHF Plus assessments were triggered for 102 people. Successful contact was made in 148 (93%) alerts (median time to contact 3 days, median call length 10 minutes). During the study, 70 out of 148 (47%) alerts identified an acute medical issue, including worsening HF for 47 (32%) alerts and non-HF issues for 33 (22%) alerts. Interventions were done in 44 cases (63% of acute issues). For those without an acute medical issue (n=78), 13 (17%) reported recent hospitalisation or intervention. Out of 148 contacts, 110 (77%) had a 30 day follow-up telephone assessment (median call length 9 minutes). There were 24 out of 30 (80%) cases with an acute medical issue and an intervention with reported improvements at 30 days. In 5 cases the person was hospitalised within 30 days of initial assessment. Average weekly workload to manage the whole population (n=415) was 96.4 minutes.
Bachtiger et al. (2021)
Study size, design and location
Prospective real-world evaluation of 435 people with Medtronic CIEDs in the UK.
Key outcomes
During the study, 87 high HFRS alerts were received and followed up for a telephone assessment. Out of 82 people, 72 (82.8%) with a high-risk score were successfully contacted. Clinical diagnoses for people with high-risk scores included isolated heart failure (18.3%), heart failure concurrent to medical problem (5.7%), alternative medical problem (10.3%) and recent hospital admission (8.0%). Telephone assessment confirmed there was no acute cause for the high-risk scores in 40.2% of people. The sensitivity for detection of worsening heart failure was 87.9% (0.77 to 0.99) and specificity was 59.4% (0.50 to 0.69).
Ahmed et al. (2020)c
Study size, design and location
Prospective evaluation of 326 people with Medtronic CIEDs in the UK.
Key outcomes
Telephone contact was made with 245 people with a high HFRS. At telephone assessment, 194 (79.1%) people reported an acute medical issue, 137 (70.6%) reported symptoms consistent with worsening heart failure needing intervention, and 57 (29.4%) had an alternative acute medical problem. 51 (26.2%) people had no apparent reason for the high-risk score. The sensitivity of CIED-based remote monitoring to identify any heart failure and non-heart failure events needing intervention was 99.5% (97.2 to 99.9%) and specificity was 65.5% (57.3 to 73.2%). Positive predictive value was 79.2%. The negative predictive value of a non-high-risk score to rule out an acute event was 98.9%. Overall accuracy of the pathway to identify an acute issue (heart failure or non-heart failure related) was 84.8%.
Sustainability
The company claims that TriageHF Plus has an environmental, health, safety and energy (EHSE) evaluation integrated into its product design, manufacturing, distribution procurement, business processes and services. It claims that this will reduce EHSE risks, improve EHSE performance of operations, improve product and packaging across their lifecycle, and promote pollution prevention and the efficient use of energy.
Recent and ongoing studies
What is the workload burden associated with using the TriageHF Plus care pathway? A Prospective TriageHF Plus evaluation. ClinicalTrials.gov identifier: NCT04177199. Status: recruiting. Indication: heart failure. Device: TriageHF Plus. Estimated study completion date: October 1, 2023. Country: United Kingdom.