Qualitative Evaluation of the Personal KinetiGraph^TM Movement Recording System in a Parkinson’s Clinic

INTRODUCTION

Parkinson’s disease (PD) is a chronic, progressive, neurodegenerative movement disorder characterized, in part, by motor symptoms of bradykinesia, muscular rigidity, and tremor. Clinical manifestation and course of disease progression varies considerably across PD patients.

While care of patients with Parkinson’s disease (PwP) is highly individualized and requires consideration of a variety of motor and non-motor symptomatology (sensory, neuropsychiatric, autonomic, etc.), pharmacologic therapeutic intervention aimed at motor symptom control is typically a key part of disease management. While clinicians rely on clinical observation during a clinic visit, typically the patient’s interim history and recall are heavily weighted when making therapeutic decisions such as medication adjustments. However, when it comes to patient reporting, motor symptom recall can be incomplete and inaccurate (e.g., mistaking dyskinesia for tremor, or being unaware of dyskinesia), and symptoms and response to therapy, which may vary in frequency and severity from day to day are often difficult to recall in any detail. As noted above, a typical clinical visit provides only a snapshot of these signs, symptoms, and variations in response to therapy over time.

Incomplete patient reporting is not a trivial issue, as an accurate evaluation of the PD disease state and response to medication over time is critical for improving diagnosis and staging, monitoring response to therapy and motor complications, improving medical treatment, enhancing surgical treatment decisions and improving rehabilitation interventions [1, 2]. More importantly, discrepancies in reporting any of the features in the day-to-day management of PD may lead to the wrong changes in medications.

New wearable sensors have the advantage of offering continuous objective measurement of patient movement during regular activities of daily living, and therefore, have the potential to provide important additional information in a more accurate way to augment clinical care of PD patients. Over the past two decades, wearable technologies have become of increasing interest due to the lack of measurement tools to objectively quantify PD motor symptoms over much longer times than the brief window of time spent in a doctor’s office [3–5]. More importantly, recent studies suggest the use of wearable sensors can inform medication choices and correlate with relevant clinical assessment scales [6, 7]. Wearable sensors have also been shown to be well tolerated by patients and the availability of personalized information can enhance patient quality of life [8]. Some of these technologies are now making the transition from research into routine clinical care.

One of these technologies is the FDA-cleared wearable Personal KinetiGraph^™ Recording System (PKG^®) Movement Recording System (PKG System), which is intended to quantify kinematics of movement disorder symptoms in conditions such as PD, including tremor, bradykinesia and dyskinesia; it includes a medication reminder, an event marker, and is intended to monitor activity associated with movement during sleep. The PKG System is indicated for use in individuals 46 to 83 years of age and was FDA cleared on August 22, 2014 (K140086). The PKG System consists of an interactive data logger (PKG Watch) that resembles a wristwatch and is worn on the right or left wrist of the body side most severely impacted by PD. It contains a battery, accelerometer, memory, an optional reminder to the subject when PD medications are due and a means for recording when PD medications are taken, as well as a capacitive sensor that detects removal of the watch from the wrist. Clinic staff configure a PKG data logger for each patient prior to its use with a simple-to-use program on a tablet computer. During patient wear time of six or more days, the PKG data logger continually collects data on patient movement and reminds the patient to acknowledge when to take their levodopa medication.

At the end of the patient wear period, the PKG data logger is returned to the clinic and data are downloaded and analyzed using the manufacturer’s (GKC) algorithm to translate raw movement data into a printable output (PKG) of the patient’s movement over the wear period. The PKG includes daily and summary scores for bradykinesia (BKS), dyskinesia (DKS), and fluctuations (FDS), data on tremor, immobility, movement during daytime, somnolence, sleep, and a summary of any periods in which the PKG data logger was off-wrist. The resulting PKG is then reviewed by a Movement Disorder Specialist (MDS) and the information used along with information gained during the routine visit to make clinical decisions.

Several published validation studies demonstrate the PKG System’s ability to quantify the kinematics of movement disorder symptoms in PwP. A validation study published in 2012 described methodology behind the PKG bradykinesia and dyskinesia algorithms and compared these outcomes with existing clinical scales. The PKG BKS was shown to closely correlate with the Unified Parkinson’s Rating Scale Part III (UPDRS III) motor score (minus tremor item) and the PKG DKS correlated with the modified Abnormal Involuntary Movement Score (AIMS) [9]. Evans, et al. demonstrated overuse of the medication acknowledgement on the PKG Watch converted to a response ratio (number of acknowledgements/number of doses) correlated in 19 of 25 subjects to impulsive-compulsive behavior (ICB) scales, suggesting an elevated PKG response ratio may be an indication of ICBs, which are often underreported in PwP [10]. The PKGs FDS was validated by Horne, et al. who describe a highly sensitive (97.1%) and selective (87.5%) score derived from BKS and DKS scores for objectively identifying PD-related fluctuations. The authors conclude the FDS promises to be a useful tool for identifying patients whose fluctuations are progressing and may require therapeutic changes [11]. Ossig, et al. compared PKG motor assessments with patient-completed home diaries. Distribution of total hours per day in all motor states measured by PKG closely reflected those assessed by PD home diaries. Consistently, the authors found a moderate correlation between calibrated PKG and diary data for total daily hours in Off and On state without dyskinesia and a strong correlation for the dyskinetic state on the group level. Authors conclude the PKG is a valuable tool to measure total motor state hours per day particularly for dyskinesias [12]. A series of published studies provided further validation of PKG use in the routine care of PwP. PKG measures of bradykinesia and dyskinesia were shown to correlate with relevant clinical scales and capture the effect of therapeutic interventions [6]. Price, et al. evaluated the clinical utility of the PKG in routine clinical care and found the PKG assisted in patient education, explaining treatment plans, and facilitating discussion around symptom management. Perhaps most importantly, in 63% of patients, the PKG identified issues that had not been reported previously [13]. Spengler, et al. reported PKG use to support deep brain stimulation (DBS) programming was feasible and may decrease time-to-DBS optimization contributing to a more effective DBS therapy and possibly fewer programming visits [14]. Additionally, a recent study using the PKG showed how a cohort of PwP with uncontrolled PD symptoms would benefit from objective assessment treatment of their PD features towards a target range [15].

The Parkinson’s Institute and Clinical Center began using the PKG System in December 2015 as an adjunct to the clinical visit history and examination. In this qualitative evaluation, we assessed the “real world” ways in which the PKG could support clinical management decisions in day-to-day patient care by physicians who were using the PKG to assess impact of this new tool on clinical decisions and patient care.

MATERIALS AND METHODS

RESULTS

This project was conducted under the supervision of Parkinson’s Institute and Clinical Center’s clinical and administrative leadership. The extent of the four-physician participation varied from 6 surveys to 46 surveys per physician among the final set of completed surveys.

During the survey period, December 2015 through July 2016, a total of 143 PKGs were ordered on 89 patients (44 patients had two or three PKGs) as part of their routine clinical evaluation and follow-up PD care visits. Of the 89 patients, 45 had one PKG, 44 had two PKGs, and 10 of the 44 patients went on to have three PKGs completed. A total of 119 Physician Surveys were completed. Four additional Physician Surveys were partially completed (having answered one of the two key questions) and for 20 patient visits where a PKG was used, a physician inadvertently did not complete the survey. Therefore, the survey completion rate was 83% (119/143). Survey completion rates were the highest for patients with only one PKG (97%) and decreased with the 2nd PKG to 64% and then to 50% at the 3rd PKG. Table 1 (Number of PKGs and Physician PKG Surveys Completed) summarizes PKG and Physician PKG Survey completion numbers by patient sub-group.

Of the 119-completed physician PKG surveys, seven were removed from analysis because they were completed prior to the first time in clinic visit, leaving a total of 112 completed surveys in 81 patients to be included in the final analysis.

Of these 112 physician surveys, 46 (41%) indicated the PKG provided additional information to the physician; however, 66 (59%) indicated the PKG provided no additional information. Upon further review of the 46 surveys, 36 surveys [36/46 (78%)] stated the information resulted in an alteration in patient care whereas, 10 surveys [10/46 (22%)] stated the PKG provided additional information but that no alteration in patient care occurred based on this information. For the overall cohort, 32% (36/112) resulted in alteration to patient care and 9% (10/112) no alteration to patient care.

Of the 36 patients the PKG most commonly yielded new and precise information on daily off time [50% of cases (18/36)]. Table 2 is a summary of the Physician PKG survey results by patient group and PKG sequence. When all patients are considered as a group, both the first and second PKG resulted in a 33% alteration in patient care.

Of the 36 total surveys reporting an alteration in patient care, the PD Symptom Fluctuator patient group had the largest proportion of surveys at 58% (21/36) followed by the First Patient Visit in Clinic patient group at 22% (8/36), the Indeterminate History patient group at 14% (5/36), and DBS/Duopa patient group at 6% (2/36).

In the PD Symptom Fluctuator patient group, changes to PD medications were the reason for all alterations in patient care. In the next section, we present three case study examples from patients in the Fluctuator patient group depicting PKG information which informed patient care. The first case demonstrates management of peak-dose dyskinesia; the second case highlights managing wearing-off without worsening existing dyskinesia; and the third case shows the patient’s subjective report of markedly wearing-off and disabling bradykinesia, is revealed as predominantly marked dyskinesia requiring further management.

Case studies

Case 1 Patient No 18

Case 1 was taking Rytary 95 mg, two capsules QID and a monoamine oxidase inhibitor (MAOI). The patient reported mild wearing-off and dyskinesia, along with gait challenges and leg dystonia. The initial PKG in January 2016 revealed symptoms more severe than reported with DKS that peaked about 90 minutes after the first dose which usually fell to near normal levels within 30–60 minutes and biphasic dyskinesia was also noted. Median DKS was 4.0 and the FDS was 11.8 indicating fluctuations were present. An anti-dyskinetic (propranolol 10 mg Qam) was added to the medication regimen during the January office visit; in March of 2016, Amantadine 100mg BID was added based on the patient phone report of continued dyskinesia. A subsequent April 2016 PKG, performed days prior to the office visit, revealed median DKS was reduced to 1.6 and FDS to 8.6. Median BKS increased slightly from 23.3 in January, to 25.7 in April; however, dyskinesia fluctuations are less pronounced. Dose-related dyskinesia is still present but reduced in severity and duration, consistent with the patient report of doing better. Figure 1 (Patient No 18 January 2016 and April 2016 PKG) depicts the two PKG printouts.

Fig. 1

Patient No 18 January 2016 and April 2016 PKG. PKG Summary Plot depicts data from recording day aligned to the time of day. It shows when reminders were given (vertical red lines), the median DKS (heavy green line) and median BKS (heavy blue line) and their corresponding 25th and 75th percentiles plotted against time of day.

Case 2 Patient No 13

Case 2 was prescribed Sinemet 25/100 one tablet five times daily, a MAOI (rasagiline 1mg), and Sinemet 50/200 CR at bedtime; the patient reported PD medication was not lasting, with greater time feeling off. Upon further questioning, it was revealed the patient had not been taking prescribed medication consistently. The initial PKG in December 2015 revealed a median BKS of 26.4 and no or limited dose response, and fluctuating symptoms with periods of moderate bradykinesia. The PKG also indicated the patient frequently acknowledged taking medication outside the prescribed dose regimen. Based on the patient report and the initial PKG, the patient was educated to the importance of consistently taking medication as prescribed. At the March 2016 visit the subsequent PKG showed improved medication compliance and the median BKS reduced to 22.5 without a substantial increase in dyskinesia. Overall, the patient was spending ∼51% of the waking day in moderate-to-severe levels of bradykinesia in December 2015, which was reduced to ∼37% in March 2016 (vs ∼25% for control patients). Figure 2 (Patient No. 13 December 2015 and March 2016 PKG) depicts the two PKG printouts.

Fig. 2

Patient No 13 December 2015 and March 2016 PKG. PKG Summary Plot depicts data from recording day aligned to the time of day. It shows when reminders were given (vertical red lines), the median DKS (heavy green line) and median BKS (heavy blue line) and their corresponding 25th and 75th percentiles plotted against time of day.

Case 3 Patient No 145

Case 3 was taking Sinemet 25/100 two tablets five times per day and Sinemet 50/200 CR ½ tablet TID and one tablet QHS. The patient reported significant motor fluctuations with wearing-off, delayed-on and challenges performing activities of daily living relative to bradykinesia. The patient was unable to say with certainty how long doses of medication last, but the patient did not report challenges with dyskinesia. Based on patient’s report, no change was made in medication and a PKG was ordered after this December 2015 office visit. Figure 3 (Patient 145 December 2015 PKG) depicts dyskinesia even though the patient did not report it. The PKG revealed a BKS score of 14.8 which is below the median for controls and approximately 68% of the day in BKI (vs ∼50% for controls), and frequent and significant fluctuations with markedly elevated median DKS of 24.8 and FDS of 19.2. The patient was contacted by phone and instructed to reduce his Sinemet 25/100 IR to one tablet five times per day and increase Sinemet 50/200 CR ½ tablet from three times to five times per day coinciding with the Sinemet IR dosing. The patient subsequently reported by phone his delayed-on and dyskinesia were both worse and was instructed by phone to return to his original medication regimen. This resulted in worsening dyskinesia and a subsequent PKG was ordered in May 2016; during his June 2016 visit, the results were reviewed and the options of DBS and Duopa were discussed to treat his significant motor fluctuations. Figure 4 (Patient 145 May 2016 PKG) depicts dyskinesia from the May 2016 PKG.

Fig. 3

Patient 145 December 2015 PKG: PKG Summary Plot depicts peak dose dyskinesia. Top Image - PKG Summary Plot: Depicts data from recording day aligned to the time of day. It shows when medication reminders were given (vertical red lines), the median DKS (heavy green line) and median BKS (heavy blue line) and their corresponding 25th and 75th percentiles plotted against time of day. Bottom Image - PKG Dose Response Curves: Depicts data from recording day aligned to the time of medication acknowledgement for each individual dose. Programmed medication doses are depicted by vertical red lines. Also illustrated are the median DKS (heavy green line) and median BKS (heavy blue line) and their corresponding 25th and 75th percentiles plotted against time of day.

Fig. 4

Patient 145 May 2016 PKG. PKG Summary Plot depicts peak dose dyskinesia. Top Image - PKG Summary Plot: Depicts data from all recording day aligned to the time of day. It shows when reminders were given (vertical red lines), the median DKS (heavy green line) and median BKS (heavy blue line) and their corresponding 25th and 75th percentiles plotted against time of day. Bottom Image - PKG Dose Response Curves: Depicts data from recording day aligned to the time of medication acknowledgement for each individual dose. Programmed medication doses are depicted by vertical red lines. Also illustrated are the median DKS (heavy green line) and median BKS (heavy blue line) and their corresponding 25th and 75th percentiles plotted against time of day.

DISCUSSION

In this qualitative evaluation, Movement Disorder Specialists in a Parkinson’s disease clinic reported the PKG provided additional information beyond that obtained during clinical consultation alone in 41% of visits, and resulted in adjusting treatment nearly a third of the time overall. The PKG most commonly yielded new and precise information on daily off time [50% of cases (18/36)], which suggests a key area where clinical consultations can derive value from the addition of continuous objective measurement.

Use of the PKG in new patient exams were found to be more beneficial when the PKG was given to the patient after the first new patient visit. Given the comprehensive data collection and information already covered at the first new patient visit, the participating physicians felt more comfortable using methods already established for the first new patient consults.

This work represents the first systematic evaluation of physician assessment on the value added by continuous objective measurement to the clinical consultation in a “real world” clinical practice in the United States. As such, these data offer new insight into the role of such technologies in the routine care of patients with PD. While many segments of PD research focus on the development of new therapies, we have learned there is also an opportunity to maximize the use of existing PD therapies in routine clinical care. Continuous objective measurement may offer a mechanism to objectively define PD motor symptoms, define treatment targets for motor function, and lead to a new paradigm for better monitoring PD patients, which may ultimately drive improved use of treatments we already have in our armamentarium as well as better outcomes for PD patients.

Integration of continuous objective measurement technology into routine clinical care required initial support staff training from the device manufacturer related to device logistics (dispensing, wear, retrieval) and physician training on clinical interpretation of the device data summary provided at the end of the wear period. Physicians also benefited from a deeper dive into the science behind how the technology works and the summary data are generally derived. Physicians completed an initial evaluation phase with report interpretation support and the learning curve for each physician varied. Clinical work flow development was required to define how the device would be ordered and provided to the patient with the aim of having the device data available at the time of the clinic visit with the patient as we found in many instances that reviewing the PKG data with the patient during the clinic visit assisted with symptom discovery. Patient training was completed in the clinic or remotely depending on the patient’s preference and familiarity with the device.

CONFLICT OF INTEREST

The authors declare there are no conflicts of interest relevant to this work.