Testing a Prediction Algorithm Into a Running Telehealth System for Patients With COPD

Can a Predictive Algorithm Used to Flag Risk of Exacerbations in a Telehealth System Strengthen the Effectiveness and Cost-effectiveness Outcomes When Monitoring Patients With COPD?

This trial will test a COPD prediction algorithm into a telehealth system from the previous Danish large-scale trial, TeleCare North (NCT01984840). The COPD prediction algorithm aims to support clinical decisions by predicting exacerbations in patients with COPD based on selected physiological parameters (blood pressure, oxygen saturation, and pulse). A prospective, parallel two-armed randomized controlled trial with approximately 200 COPD participants will be conducted.

Several studies call for research investigating telehealth' ability to predict exacerbations. Use of clinical prediction tools might be promising to improve telehealth services related to prediction of exacerbations and to support decision-making. However, more research is needed to further develop, test, and validate prediction algorithms to ensure that these algorithms improve clinical outcomes before they are widely implemented in practice. This trial seeks to demonstrate that through telehealth, the implementation of a COPD prediction algorithm might have potential to support early detection of exacerbations. The COPD prediction algorithm might initiate timely treatment, which can potentially led to improvement in COPD patients' health and fewer hospitalizations. The primary outcome is the number of exacerbations defined as an all-cause acute hospitalization from baseline to follow-up in both groups. The trial hypothesizes that integrating a COPD prediction algorithm into the telehealth system will lead to a significantly lower number of exacerbations through early identification and timely preventive treatment. The primary outcome will be statistically analyzed, and the hypothesis will be tested between groups. All participants are familiar with the telehealth system in advance. In addition to the participants' usual monitored measurements, they are asked to measure their oxygen saturation twice a week during the trial period. The participants will receive the questionnaires; EuroQol-5D-5L, Short-Form 12 item, version 2 Health Survey, The European Health Literacy Survey Questionnaire, The Danish Test of Functional Health Literacy in Adults, Danish Telehealth Usability Questionnaire, and a questionnaire containing selected demographic characteristics at baseline and at 6-month follow-up.

chronic obstructive pulmonary disease, Denmark, acute episodes, machine learning, telehealth, telemonitoring, RCT, decision support systems, predictive analytics

The study is single-blinded, in which the participants do not know the treatment group that they have been assigned, i.e. the participants do not know if the COPD prediction algorithm runs in the background of their profile or not. The specialized COPD community nurses cannot be blinded, as they need to be trained in the use of the COPD prediction algorithm.

The specialized COPD community nurses, who are responsible for patients in the intervention group, will experience an extra alarm option in the telehealth system. The COPD prediction algorithm has been implemented by the Danish Company, OpenTeleHealth, into their commercially available telehealth system, Telekit, and thus, the COPD prediction algorithm is approved for clinical use as a part of the existing telehealth system´s CE marking (class I and IIa).

The specialized COPD community nurses responsible for patients in the control group will only experience the usual alarms that are activated in the telehealth system, named Telekit, and are based on low or high values of vital signs. The specialized COPD community nurses are not instructed to act differently compared to how they act normally. This involves that the specialized COPD community nurses monitor as usual and respond to divergent data as usual. The participants in the control group receive the usual practice, which includes the general offer of the telehealth intervention. The participants in the control group are instructed to do exactly the same procedures as the participants in the intervention group. The specialized COPD community nurses, who are responsible for patients in the control group continue to monitor the participants as usual, but are informed that more oxygen saturation measurements will be present for the included participants.

Participants in the intervention group will receive the general offer of the telehealth intervention including the telehealth system with the implemented COPD prediction algorithm. The participants are asked to measure their vital signs and respond to COPD related questions as usual. "As usual" refers to fixed days in the week, either Monday or Thursday, where the participants must measure their vital signs. In addition, the participants must weekly measure two oxygen saturation measurements drawn from the fingertip pulse oximeter.

The primary outcome is the number of exacerbations defined as an all-cause acute hospitalization from baseline to follow-up in both groups. The trial hypothesizes that integrating a COPD prediction algorithm into the telehealth system will lead to a significantly lower number of exacerbations through early identification and timely preventive treatment

To compare the change in health-related quality of life (SF-12v2) at the individual level from baseline to follow-up at 6 months.

The change in health-related quality of life (HRQoL) using SF-12v2 at the individual level from baseline to follow-up at 6 months. The trial hypothesizes that the difference in HRQoL from baseline to follow-up in both groups decrease since the participants have lived six months longer with COPD. However, it is expected that the decrease in HRQoL will be lower for the intervention group compared to the control group

To compare the change in health-related quality of life (EQ5D-5L) at the individual level from baseline to follow-up at 6 months.

The change in health-related quality of life (HRQoL) using EQ-5D-5L at the individual level from baseline to follow-up at 6 months. The trial hypothesizes that the difference in HRQoL from baseline to follow-up in both groups decrease since the participants have lived six months longer with COPD. However, it is expected that the decrease in HRQoL will be lower for the intervention group compared to the control group

To compare the ICER (EQ-5D-5L) measured as the cost per quality adjusted life years (QALY) at the individual level from baseline to follow up at 6 months

The incremental cost-effectiveness ratio or ICER measured as the total cost per quality adjusted life years (QALY) gained for the cost-categories included in the study from baseline to follow up at six months. It is hypothesized that the cost of hospital contacts will decrease, but it is unknown whether this cost is offset by an increase in other cost-categories such as community care

To assess the participants' health literacy level at baseline using the HLS-EU-Q16, supported by further assessment with the Danish TOFHLA during the trial period to examine whether the effect of the COPD prediction algorithm is similar in patients with COPD, regardless of health literacy level

To examine whether the specialized COPD community nurse's estimate of the individual participant's level of health literacy influences the effect of the COPD prediction algorithm

Evaluation of the specialized COPD community nurses' experiences with the usability of the COPD prediction algorithm

To evaluate the specialized COPD community nurses' experiences with the usability of the COPD prediction algorithm using interviews

To evaluate the participants' experiences with the usability of the telehealth system after trial completion using the questionnaire D-TUQ

To evaluate the participants' experiences with data ethical aspects after trial completion using qualitative research interviews

The trial population consists of patients with COPD who already use the telehealth system. Inclusion criteria: Men and women >18 years Diagnosis of COPD Fixed residence in Aalborg Municipality. Exclusion criteria: Unable to monitor vital signs Unable to complete study questionnaires

Kronborg T, Mark L, Cichosz SL, Secher PH, Hejlesen O. Population exacerbation incidence contains predictive information of acute exacerbations in patients with chronic obstructive pulmonary disease in telecare. Int J Med Inform. 2018 Mar;111:72-76. doi: 10.1016/j.ijmedinf.2017.12.026. Epub 2017 Dec 29.

Udsen FW, Lilholt PH, Hejlesen O, Ehlers LH. Effectiveness and cost-effectiveness of telehealthcare for chronic obstructive pulmonary disease: study protocol for a cluster randomized controlled trial. Trials. 2014 May 21;15:178. doi: 10.1186/1745-6215-15-178.

Lilholt PH, Witt Udsen F, Ehlers L, Hejlesen OK. Telehealthcare for patients suffering from chronic obstructive pulmonary disease: effects on health-related quality of life: results from the Danish 'TeleCare North' cluster-randomised trial. BMJ Open. 2017 May 9;7(5):e014587. doi: 10.1136/bmjopen-2016-014587.

Lilholt PH, Haesum LK, Hejlesen OK. Exploring User Experience of a Telehealth System for the Danish TeleCare North Trial. Stud Health Technol Inform. 2015;210:301-5.

Witt Udsen F, Lilholt PH, Hejlesen O, Ehlers L. Cost-effectiveness of telehealthcare to patients with chronic obstructive pulmonary disease: results from the Danish 'TeleCare North' cluster-randomised trial. BMJ Open. 2017 May 17;7(5):e014616. doi: 10.1136/bmjopen-2016-014616.

Secher PH, Hangaard S, Kronborg T, Haesum LKE, Udsen FW, Hejlesen O, Bender C. Clinical implementation of an algorithm for predicting exacerbations in patients with COPD in telemonitoring: a study protocol for a single-blinded randomized controlled trial. Trials. 2022 Apr 26;23(1):356. doi: 10.1186/s13063-022-06292-y.