Table 2.
Main concerns and questions of the CHMP for the applicants.
| Main recommendations |
|---|
| A. Relevance of the (novel) outcome measure as assessed by the remote monitoring technology |
| Is the outcome measure relevant for the disease of interest? Does the device measure all aspects or symptoms of a specific disease? Does the device measure all aspects of a specific function? To what extent is the outcome measure only influenced by the disease activity of interest? How does the outcome measure relate to the CHMP guideline of the corresponding disease? In case of sensor data: to what extent is the body part to which the sensor is attached reflective of the symptoms of the disease or condition of interest? |
| B. Validation of the novel outcome measure as assessed by the remote monitoring technology |
| Is the outcome measure correlated with hard endpoints? (morbidity/mortality) Is the outcome measure correlated with relevant outcome measures for patients? [Quality of life (QOL) and patient-reported outcome measures (PROMs)] Is the outcome measure correlated with the established/golden standard clinical tests and/or PROMs of the disease of interest? In case several established outcome measures exist that measure different symptoms or aspects of a disease: to what extent does the novel outcome correlate with all these different endpoints? (NB relevance of an outcome measure can depend on the intended treatment effect) Is a change in the novel endpoint correlated with a change in final endpoints or (other) outcomes that matter to patients? (QOL or PROMs) What is the external validity? In how many patients/persons has the device been tested? Is the minimal clinically important change determined? If yes: is this studied prospectively? Is the minimal clinically important change determined for the different diseases, subgroups and clinical stages of interest? |
| C. Precision, accuracy, sensitivity, and specificity |
| To what extent does the novel endpoint predict the established endpoint(s) with enough precision? What is the internal validity? (same value in stable patients) Are there systematic errors in measurements in specific subgroups? (e.g., overestimation of walking speed in more severe multiple sclerosis patients due to ataxia) In case several outcome measures are available for a device: what outcome measure is chosen and why? What is the effect of outliers and was this taken into account? To what extent can the device check what the activity of the patient is during the measurement? If relevant: how is this issue handled? Is the outcome measure sensitive enough to distinguish relevant subgroups? Technical correctness: to what extent is the device capable of measuring a change that is clinically relevant? |
| D. Compliance and handling of missing values |
| What is the compliance? Is the compliance stable during follow-up? Is selective non-compliance an issue? (e.g., a patient does not perform tests or wears a device during periods of increased symptomatology) What measures are taken to prevent (selective) missing values? Is compliance actively stimulated? (e.g., by the use of alarms, phone calls, etc.) How are missing data handled? In case of an active test: is the test performed every day at the same time? Is a medication tracker used? (this can be of relevance in case of clear on/off states such as in Parkinson's disease) Wat is the tolerability and acceptability of the technology for patients? |
| E. Sampling interval |
| Is the sampling interval long enough to take day-to-day variation into account? Is the sampling interval per measurement short enough that no clinical change is expected and optimal compliance is expected? How is the sampling interval determined? |
| F. Privacy and data handling |
| How are data anonymized and protected? Who has access to the data? Is a risk analysis performed to guarantee optimal data security? |