• Parky operates under the following constraints:
◦ The patient should wear Apple Watch on the most affected arm.
◦ Dyskinetic symptom tracking should be considered only for patients with chorea on the affected arm, either self-reported or diagnosed by a clinician.
◦ Parky explicitly tracks resting tremor; it does not track action tremor or postural tremor, and it may not track finger tremor.
◦ Parky does not explicitly track dystonia or bradykinesia.
◦ The results may include false positives and false negatives.The patient’s activity, watch band fit, and concomitant conditions (such as restless legs syndrome and non-Parkinsonian tremor) can affect the quality of the results.
◦ Parky only explicitly measures symptoms from the wrist while wearing Apple Watch. However, Apple Watch may sense symptoms transmitted through the body from other affected body parts, possibly resulting in misleading or false metrics.
◦ Parky does not operate on unsupported devices.
◦ Parky only operates with MM4PD software version 1.0.0, developed by Apple. Any version change on MM4PD is automatically detected, and Parky operation will be suspended until a new validation is conducted.
• Parky is not advised to be used on other Parkinsonian syndromes other than idiopathic Parkinson’s Disease, such as progressive supranuclear palsy (PSP), multiple system atrophy (MSA), and corticobasal degeneration (CBD).
• Parky cannot collect data when Apple Watch is turned off and cannot transmit data when iPhone is turned off.
• Without sufficient recording times, Parky will be unable to generate symptom reports. The reports are generated using the following approach:
◦ 15-minute periods comprising a 24-hour day are organized according to the local time of the patient.
◦ Windows with less than 50% of available data are discarded.
◦ Tremor percentages are averaged over all days with data. Dyskinesia likely presence is displayed as the median.
◦ Only windows with data from at least 5 days or more than 20% of the total period are shown for the weekly and monthly reports.
• The symptom reports should not be assessed solely for disease management without evaluating non-motor symptoms, patient-reported insights, and in-clinic visits. Data provided by Parky should be considered together with the patient’s clinical history, motor and non-motor complications, and other clinical findings, as well as recording times.
• Medication schedules are subject to the correctness of entered data by the patient. If the patient is not competent enough, it is strongly advised to provide support in-clinic settings.
• Medication compliance information is subject to patient adherence for responding to med reminder notifications. Non-responded notifications or misuse can generate incorrect numbers for medication compliance. To lower this risk, it is strongly advised that the health care providers clearly underline the importance of reminder responses to the patients.
• Steps information is subject to patient adherence to Apple Watch and iPhone use. Parky cannot collect step information when the patient is not using any of these devices.
• If the patient decides to stop using Parky, they may delete their clinician information from the app. In that case, the patient is considered a dropout from treatment, and the activation code is immediately switched to an invalid state. Patients will not be able to view or share their symptom reports with an invalid activation code; thus, the clinician will not be able to receive any symptom reports. A dropout patient notification will be sent to the prescriber healthcare provider immediately. A new prescription and activation code will be required to restart using Parky.
