Innovations in Physiologic and Patient-Reported Data Capture: Implications for Streamlining Data Collection and Leveraging Access to Real-time Data Sonya Eremenco MA , United BioSource Corporation, Bethesda, MD, USA;  Wilhelm Muehlhausen DVM , Cardinal Health Germany 234 GmbH, Hoechberg, Germany;  Lionel Tarassenko FREng , University of Oxford, Institute of Biomedical Engineering, Oxford, United Kingdom;  Jill Platko, PhD  PHT Corporation, Boston, MA, USA.  ISPOR International Conference, May 18, 2009
Workshop Outline Introductions Overview of topic and technology Case example: Asthma study Case example: Diabetes study Potential for use in adverse event safety monitoring Checklist development: interactive process
Workshop Purpose Clinical trial endpoints involve patient collection of physiologic and patient-reported outcome (PRO) data  Technology enables remote collection and transmission Address the opportunities and challenges of parallel biometric-PRO data collection Empirical evidence of product value through clinical studies conducted pre and post-launch and  Ways to leverage real-time data access
Historical approach to physiologic data capture Asthma and diabetes patients monitor their conditions regularly using peak flow meters and glucometers.  Until recently, these data were collected from patients separately from PRO data during clinical trials  Increasing respondent burden  Increasing risk of error associated with patients or sites transposing and manually entering data.
Issues with manual entry of patient data  Discrepancies with paper diaries   J E Broderick et al.  Annals of Behavioral Medicine,  2003:   26:139-148. RN Jamison et al.  Pain  2001;91:277–85. TM Palermo et al.   Pain  2004, 7:213–219. A Stone et al,  BMJ  2002; 324:1193-4 AA Stone et al.  Control Clin Trials  2003; 24:182–99.
Issues with manual entry of patient data  Discrepancies with glucometer entries R. Mazze et al, The Amer. J. of Medicine,77,1984 Langer O, Mazze RS. Am J Obstet Gynecol. 1986 Sep;155(3):635-7. Williams CD et al.  Diabet Med. 1988 Jul-Aug;5(5):459-62. Ziegler O et al.,  Diabetes Care. 1989 Mar;12(3):184-8. Overreporting (addition of phantom values in logbook) Underreporting (omission of SMBG measurements from logbook) Alteration of values Improvement with electronic capture Laffel LM et al. Continued use of an integrated meter with electronic logbook maintains improvements in glycemic control beyond a randomized, controlled trial. Diabetes Technol Ther. 2007 Jun;9(3):254-64.
Issues with manual entry of patient data Peak Flow meters:  Errors or falsification P. Verschelden et al, Eur Respir J,9,1996 A W A Kamps et al., Peak flow diaries in childhood asthma are unreliable. Thorax 2001;56:180-182 (March)   Impact of electronic H. Reddel, et al., Analysis of adherence to peak flow monitoring when recording of data is electronic.  BMJ  2002;324:146-147 ( 19 January )
Manual entry problematic Regardless of whether physiologic data are entered in paper logs or diaries or in electronic diaries prone to errors: Overreporting Underreporting Alteration of values Unless the data are captured and transferred electronically (simultaneously), difficult to retrospectively link the data points for statistical analysis Times are off Different time settings No session to connect data
Convergence in technology Development of biometric devices  Use of electronic PRO administration (ePRO) via PDA/handheld devices Development of integrated devices Evolution of data transmission technology Greater technologic sophistication of consumers  Result: ability to simultaneously capture and transmit physiologic and PRO parameters in clinical studies
Other Therapeutic Areas Hypertension: blood pressure Oral Anticoagulation Therapy: prothrombin time meters Sleep COPD Obesity
Range of devices Glucometers Peak Flow meters Blood pressure meters Puls Oximeters Weight scales Activity monitors, accelerometers Sleep movement monitors Integrated vs. separate
Range of transmission methods Wireless Bluetooth  PDA Mobile phone Modems Infrared (radio frequency) ZigBee - Personal Area Network GSM/GPRS built-In Wired USB Serial
Confidential   Cardinal Health Research Services „CHRS“ ISPOR Orlando   Willie Muehlhausen May, 2009   Senior Product Manager
Case studies Persistent Asthma 8 weeks treatment 24 countries (6 continents) 1900 randomized subjects Age 12 and older Data transmission at Site (every 2 weeks) Physiologic data Peak Expiratory Flow (PEF) PRO data Rescue medication Asthma Symptoms Scores (AM & PM)
Case studies
Case studies Data Access in „Realtime“ Exclusion criteria Less than 4 / 7 days with data during Screening Early withdrawal criteria More than x number of Rescue Med use
Case studies Chronic Asthma 12 weeks treatment US only 295 randomized subjects Adults Data transmission at Site (every 4 weeks) Physiologic data Peak Expiratory Flow (PEF) PRO data Rescue medication Asthma Symptoms Scores (AM & PM)
Case studies
Case studies Screening Monitor Compliance during Screening to withdraw non-compliant subjects  Parallel Capture 99.9% of subjects who answer the questions will also do the PEF at the same time
Diabetes Case Study using  mPRO™ Data Collection ISPOR May 2009 Lionel Tarassenko FREng,  University of Oxford, Institute of Biomedical Engineering,  Oxford, United Kingdom t+Clinical
Technology for self-management? Wilson  et al.  (BMJ, 2005):   “ The evidence backing the use of disease-specific self-management programmes like diabetes is strong. The challenge is how to move to a  programme that can support the many millions of patients who might benefit . ” Focus on mobile phone: Equality of care  –  90% of UK population owns a mobile phone Real-time feedback, with two-way information flow Communication with remote carer based on shared data Economic model based on reduction in unplanned hospital admissions makes mobile phone solution a financially viable proposition
The  Solution t+ Medical  Nursing Team Prioritisation of patients Clinical management tool Red alert responses Compliance monitoring Education delivery Medicines optimisation Admissions avoidance programmes Mobile health tool Regular support from t+ telehealth nurse (based on real-time data) Interactive to promote self management Carer Alerts Colour coded feedback Healthcare practitioner
Blood Glucose Measurements Using a glucometer to test blood sugar levels, the readings can then be downloaded wirelessly over Bluetooth to the mobile phone and  submitted with the diary information to the main server.
Example patient selection All adult or transitional (juvenile to adult services) patients with an HbA1c > 7.4%, who meet one of the following criteria: Type 1 diabetes Type 2 diabetes on insulin Type 2 diabetes on oral hypoglycaemic agents, testing at least twice a week
t+ diabetes phone diary
t+ diabetes phone diary
t+ diabetes phone diary
t+ diabetes phone diary
Diabetes home page Clinician
Patients charts Clinician
Patients charts Clinician
Oxford Diabetes Type 1 clinical trial 9-month Randomised Controlled Trial with patients from Young Adult Clinic  Inclusion criteria: Type 1 diabetes, aged between 18 and 30 Twice daily or basal bolus insulin therapy Poor glycaemic control (HbA 1C  between 8 and 11%)  Aim to detect a difference of 0.7% in HbA 1C  based on baseline mean value of 9% Division of Public Health and Primary Care University of Oxford Principal Investigators: Prof. L. Tarassenko Prof. A. Neil - Dr A. Farmer
Telehealth nurse contacts 601 phone calls initiated by telehealth nurses An average of 13 calls per patient (1.5 calls per month) Duration of phone calls was 7 min 9 sec (Standard Deviation 4min 15 sec) Interaction with patients on basis of  shared data
Results  –  patient compliance “ A randomised controlled trial of the effect of real-time telemedicine support on glycemic control in young adults with type 1 diabetes” DIABETES CARE, VOLUME 28, NUMBER 11, NOVEMBER 2005
Changes in HbA 1c  over 9 months HbA 1c  (%)
Use of t+ diabetes for insulin titration Automatic alerts for hypoglycaemia Time plots of BG readings + insulin doses (last two weeks plus trend screen) available on patient ’ s web page Direct access to patient on mobile phone
Use of t+ diabetes for insulin titration (Oxfordshire GP Practices)
Insulin titration results
Real-time data collection and data access in clinical trial safety monitoring Jill V. Platko PhD Scientific Advisor 617-973-3252 [email_address] 18 May 2009
Discussion Topics Safety Monitoring (Monitoring for potential Adverse Events) Case example: Insomnia Trial-Suicidal Ideation Case example: Asthma Trial-Worsening Condition More than just Clinical Trials: Disease Management
Adverse events and paper Safety Monitoring General (non-study specific) issues  Indication or drug class specific issues Exam at study visits Specific Question(naires) at study visits Take home diary
Suicide Ideation
Suicide Ideation
Memory of Moods Paisecki et.al., Psychol Assess. 2007 Mar;19(1):25-43. eDiary 4-Week Recall
Case Study: Insomnia Trial Study details International trial to treat insomnia in subjects with Major Depressive Disorder Site Visits spaced up to 4 weeks apart Quick Inventory of Depressive Symptomatology (QIDS-SR16) completed weekly at home. Data available in real-time for review by site coordinators and sponsors
Suicide Ideation
eMail Alert From: studySupport@phtcorp.com [ mailto:studySupport@phtcorp.com ] Sent: Friday, January 09, 2009 5:17 AM To: undisclosed-recipients Subject: Urgent –  study participant Suicidal Ideation Alert DO NOT REPLY TO THIS MESSAGE!   Dear Investigator, A Subject answered the QIDS question 12 in a manner that indicates suicidal ideation. Please contact the subject immediately and determine if the subject remains eligible for the trial. Please contact Dr. Smith, M.D. at 973-986-3456 if you have any questions. Dr. Smith will be contacting you within 24-48 hours to discuss the subject. Sincerely, The QIDS Alert System
Data Summary
Asthma Trials
Peak Flow Meter
Recording PEF Values We conclude that: 1) compliance with daily peak expiratory flow assessments is generally poor (…) and 2) a substantial percentage of values (22%) is invented. Conclusions- Peak flow diaries kept by asthmatic children are unreliable.  Electronic peak flow meters should be used if peak flow monitoring is required in children with asthma. Compliance with and accuracy of daily self-assessment of peak expiratory flows (PEF) in asthmatic subjects over a three month period. Verschelden P, Cartier A, L'Archevêque J, Trudeau C, Malo JL. Eur Respir J. 1996 May;9(5):880-5 Peak flow diaries in childhood asthma are unreliable. Kamps AW, Roorda RJ, Brand PL. Thorax. 2001 Mar;56(3):180-2.
eSense Case Study: Global Asthma Trial Study details International trial with > 500 subjects in 15 countries using portable electronic PEF meter Subjects take morning and evening PEF values using an eSense PiKo meter by Ferraris Respiratory   Data is available in real-time for review by site coordinators and sponsors
PEF Values Alerts The following specifications should generate a patient diary alert: Peak Flow measurement less than 50% of Subject’s personal best Explanation: The LP will update and store the Subject’s personal best Peak Flow (morning or evening) on an on-going basis, then compare every new entry to this stored value. If the new Peak Flow measurement is less than to 50% of this stored value, the LP will alert. Alert: “You have indicated a large drop in Peak Flow. Please contact your Study Coordinator.” Example   Day 5: Subject enters 300 (LP stores personal best) Day 50: Subject enters 325 (LP updates personal best) Day 149: Subject enters 155 – LP alerts Diary Alerts
Case Study: Disease Management Challenge:  Reduce hospitalizations and deaths due to COPD exacerbations; improve patient lung function and quality of life. Methods:  Subjects use the LogPad daily to answer questions, which are scored in real time; if the score crosses certain thresholds, patients are instructed to contact the hospital’s pulmonary call center for instruction.
Report for Patient Monitoring
Clinician Reaction “ This technology, which could easily be used with other diseases, has a truly great impact on patients’ quality of life and the disease cost and burden on society,” Dr. Wissam Chatila  Temple Lung Center
Patient Reaction "With the (LogPad) you have a daily routine, seven days a week, so you can't miss it. You can't go wrong … This here keeps me out of the hospital as much as possible. That's what I love about it."  Edward Goldwire COPD patient
Any Questions
Discussion 1 What factors determine which studies would benefit from combining physiologic and PRO data collection?
Discussion 2 What barriers to implementing parallel data capture do you see in your company?
Discussion 3 What strategies can be used to overcome the barriers to adoption?

Ispor2009: Innovations In Physiologic And Pro Data Capture

  • 1.
    Innovations in Physiologicand Patient-Reported Data Capture: Implications for Streamlining Data Collection and Leveraging Access to Real-time Data Sonya Eremenco MA , United BioSource Corporation, Bethesda, MD, USA; Wilhelm Muehlhausen DVM , Cardinal Health Germany 234 GmbH, Hoechberg, Germany; Lionel Tarassenko FREng , University of Oxford, Institute of Biomedical Engineering, Oxford, United Kingdom; Jill Platko, PhD PHT Corporation, Boston, MA, USA. ISPOR International Conference, May 18, 2009
  • 2.
    Workshop Outline IntroductionsOverview of topic and technology Case example: Asthma study Case example: Diabetes study Potential for use in adverse event safety monitoring Checklist development: interactive process
  • 3.
    Workshop Purpose Clinicaltrial endpoints involve patient collection of physiologic and patient-reported outcome (PRO) data Technology enables remote collection and transmission Address the opportunities and challenges of parallel biometric-PRO data collection Empirical evidence of product value through clinical studies conducted pre and post-launch and Ways to leverage real-time data access
  • 4.
    Historical approach tophysiologic data capture Asthma and diabetes patients monitor their conditions regularly using peak flow meters and glucometers. Until recently, these data were collected from patients separately from PRO data during clinical trials Increasing respondent burden Increasing risk of error associated with patients or sites transposing and manually entering data.
  • 5.
    Issues with manualentry of patient data Discrepancies with paper diaries J E Broderick et al. Annals of Behavioral Medicine, 2003: 26:139-148. RN Jamison et al. Pain 2001;91:277–85. TM Palermo et al. Pain 2004, 7:213–219. A Stone et al, BMJ 2002; 324:1193-4 AA Stone et al. Control Clin Trials 2003; 24:182–99.
  • 6.
    Issues with manualentry of patient data Discrepancies with glucometer entries R. Mazze et al, The Amer. J. of Medicine,77,1984 Langer O, Mazze RS. Am J Obstet Gynecol. 1986 Sep;155(3):635-7. Williams CD et al. Diabet Med. 1988 Jul-Aug;5(5):459-62. Ziegler O et al., Diabetes Care. 1989 Mar;12(3):184-8. Overreporting (addition of phantom values in logbook) Underreporting (omission of SMBG measurements from logbook) Alteration of values Improvement with electronic capture Laffel LM et al. Continued use of an integrated meter with electronic logbook maintains improvements in glycemic control beyond a randomized, controlled trial. Diabetes Technol Ther. 2007 Jun;9(3):254-64.
  • 7.
    Issues with manualentry of patient data Peak Flow meters: Errors or falsification P. Verschelden et al, Eur Respir J,9,1996 A W A Kamps et al., Peak flow diaries in childhood asthma are unreliable. Thorax 2001;56:180-182 (March) Impact of electronic H. Reddel, et al., Analysis of adherence to peak flow monitoring when recording of data is electronic. BMJ 2002;324:146-147 ( 19 January )
  • 8.
    Manual entry problematicRegardless of whether physiologic data are entered in paper logs or diaries or in electronic diaries prone to errors: Overreporting Underreporting Alteration of values Unless the data are captured and transferred electronically (simultaneously), difficult to retrospectively link the data points for statistical analysis Times are off Different time settings No session to connect data
  • 9.
    Convergence in technologyDevelopment of biometric devices Use of electronic PRO administration (ePRO) via PDA/handheld devices Development of integrated devices Evolution of data transmission technology Greater technologic sophistication of consumers Result: ability to simultaneously capture and transmit physiologic and PRO parameters in clinical studies
  • 10.
    Other Therapeutic AreasHypertension: blood pressure Oral Anticoagulation Therapy: prothrombin time meters Sleep COPD Obesity
  • 11.
    Range of devicesGlucometers Peak Flow meters Blood pressure meters Puls Oximeters Weight scales Activity monitors, accelerometers Sleep movement monitors Integrated vs. separate
  • 12.
    Range of transmissionmethods Wireless Bluetooth PDA Mobile phone Modems Infrared (radio frequency) ZigBee - Personal Area Network GSM/GPRS built-In Wired USB Serial
  • 13.
    Confidential Cardinal Health Research Services „CHRS“ ISPOR Orlando Willie Muehlhausen May, 2009 Senior Product Manager
  • 14.
    Case studies PersistentAsthma 8 weeks treatment 24 countries (6 continents) 1900 randomized subjects Age 12 and older Data transmission at Site (every 2 weeks) Physiologic data Peak Expiratory Flow (PEF) PRO data Rescue medication Asthma Symptoms Scores (AM & PM)
  • 15.
  • 16.
    Case studies DataAccess in „Realtime“ Exclusion criteria Less than 4 / 7 days with data during Screening Early withdrawal criteria More than x number of Rescue Med use
  • 17.
    Case studies ChronicAsthma 12 weeks treatment US only 295 randomized subjects Adults Data transmission at Site (every 4 weeks) Physiologic data Peak Expiratory Flow (PEF) PRO data Rescue medication Asthma Symptoms Scores (AM & PM)
  • 18.
  • 19.
    Case studies ScreeningMonitor Compliance during Screening to withdraw non-compliant subjects Parallel Capture 99.9% of subjects who answer the questions will also do the PEF at the same time
  • 20.
    Diabetes Case Studyusing mPRO™ Data Collection ISPOR May 2009 Lionel Tarassenko FREng, University of Oxford, Institute of Biomedical Engineering, Oxford, United Kingdom t+Clinical
  • 21.
    Technology for self-management?Wilson et al. (BMJ, 2005): “ The evidence backing the use of disease-specific self-management programmes like diabetes is strong. The challenge is how to move to a programme that can support the many millions of patients who might benefit . ” Focus on mobile phone: Equality of care – 90% of UK population owns a mobile phone Real-time feedback, with two-way information flow Communication with remote carer based on shared data Economic model based on reduction in unplanned hospital admissions makes mobile phone solution a financially viable proposition
  • 22.
    The Solutiont+ Medical Nursing Team Prioritisation of patients Clinical management tool Red alert responses Compliance monitoring Education delivery Medicines optimisation Admissions avoidance programmes Mobile health tool Regular support from t+ telehealth nurse (based on real-time data) Interactive to promote self management Carer Alerts Colour coded feedback Healthcare practitioner
  • 23.
    Blood Glucose MeasurementsUsing a glucometer to test blood sugar levels, the readings can then be downloaded wirelessly over Bluetooth to the mobile phone and submitted with the diary information to the main server.
  • 24.
    Example patient selectionAll adult or transitional (juvenile to adult services) patients with an HbA1c > 7.4%, who meet one of the following criteria: Type 1 diabetes Type 2 diabetes on insulin Type 2 diabetes on oral hypoglycaemic agents, testing at least twice a week
  • 25.
  • 26.
  • 27.
  • 28.
  • 29.
  • 30.
  • 31.
  • 32.
    Oxford Diabetes Type1 clinical trial 9-month Randomised Controlled Trial with patients from Young Adult Clinic Inclusion criteria: Type 1 diabetes, aged between 18 and 30 Twice daily or basal bolus insulin therapy Poor glycaemic control (HbA 1C between 8 and 11%) Aim to detect a difference of 0.7% in HbA 1C based on baseline mean value of 9% Division of Public Health and Primary Care University of Oxford Principal Investigators: Prof. L. Tarassenko Prof. A. Neil - Dr A. Farmer
  • 33.
    Telehealth nurse contacts601 phone calls initiated by telehealth nurses An average of 13 calls per patient (1.5 calls per month) Duration of phone calls was 7 min 9 sec (Standard Deviation 4min 15 sec) Interaction with patients on basis of shared data
  • 34.
    Results – patient compliance “ A randomised controlled trial of the effect of real-time telemedicine support on glycemic control in young adults with type 1 diabetes” DIABETES CARE, VOLUME 28, NUMBER 11, NOVEMBER 2005
  • 35.
    Changes in HbA1c over 9 months HbA 1c (%)
  • 36.
    Use of t+diabetes for insulin titration Automatic alerts for hypoglycaemia Time plots of BG readings + insulin doses (last two weeks plus trend screen) available on patient ’ s web page Direct access to patient on mobile phone
  • 37.
    Use of t+diabetes for insulin titration (Oxfordshire GP Practices)
  • 38.
  • 39.
    Real-time data collectionand data access in clinical trial safety monitoring Jill V. Platko PhD Scientific Advisor 617-973-3252 [email_address] 18 May 2009
  • 40.
    Discussion Topics SafetyMonitoring (Monitoring for potential Adverse Events) Case example: Insomnia Trial-Suicidal Ideation Case example: Asthma Trial-Worsening Condition More than just Clinical Trials: Disease Management
  • 41.
    Adverse events andpaper Safety Monitoring General (non-study specific) issues Indication or drug class specific issues Exam at study visits Specific Question(naires) at study visits Take home diary
  • 42.
  • 43.
  • 44.
    Memory of MoodsPaisecki et.al., Psychol Assess. 2007 Mar;19(1):25-43. eDiary 4-Week Recall
  • 45.
    Case Study: InsomniaTrial Study details International trial to treat insomnia in subjects with Major Depressive Disorder Site Visits spaced up to 4 weeks apart Quick Inventory of Depressive Symptomatology (QIDS-SR16) completed weekly at home. Data available in real-time for review by site coordinators and sponsors
  • 46.
  • 47.
    eMail Alert From:studySupport@phtcorp.com [ mailto:studySupport@phtcorp.com ] Sent: Friday, January 09, 2009 5:17 AM To: undisclosed-recipients Subject: Urgent – study participant Suicidal Ideation Alert DO NOT REPLY TO THIS MESSAGE! Dear Investigator, A Subject answered the QIDS question 12 in a manner that indicates suicidal ideation. Please contact the subject immediately and determine if the subject remains eligible for the trial. Please contact Dr. Smith, M.D. at 973-986-3456 if you have any questions. Dr. Smith will be contacting you within 24-48 hours to discuss the subject. Sincerely, The QIDS Alert System
  • 48.
  • 49.
  • 50.
  • 51.
    Recording PEF ValuesWe conclude that: 1) compliance with daily peak expiratory flow assessments is generally poor (…) and 2) a substantial percentage of values (22%) is invented. Conclusions- Peak flow diaries kept by asthmatic children are unreliable. Electronic peak flow meters should be used if peak flow monitoring is required in children with asthma. Compliance with and accuracy of daily self-assessment of peak expiratory flows (PEF) in asthmatic subjects over a three month period. Verschelden P, Cartier A, L'Archevêque J, Trudeau C, Malo JL. Eur Respir J. 1996 May;9(5):880-5 Peak flow diaries in childhood asthma are unreliable. Kamps AW, Roorda RJ, Brand PL. Thorax. 2001 Mar;56(3):180-2.
  • 52.
    eSense Case Study:Global Asthma Trial Study details International trial with > 500 subjects in 15 countries using portable electronic PEF meter Subjects take morning and evening PEF values using an eSense PiKo meter by Ferraris Respiratory Data is available in real-time for review by site coordinators and sponsors
  • 53.
    PEF Values AlertsThe following specifications should generate a patient diary alert: Peak Flow measurement less than 50% of Subject’s personal best Explanation: The LP will update and store the Subject’s personal best Peak Flow (morning or evening) on an on-going basis, then compare every new entry to this stored value. If the new Peak Flow measurement is less than to 50% of this stored value, the LP will alert. Alert: “You have indicated a large drop in Peak Flow. Please contact your Study Coordinator.” Example Day 5: Subject enters 300 (LP stores personal best) Day 50: Subject enters 325 (LP updates personal best) Day 149: Subject enters 155 – LP alerts Diary Alerts
  • 54.
    Case Study: DiseaseManagement Challenge: Reduce hospitalizations and deaths due to COPD exacerbations; improve patient lung function and quality of life. Methods: Subjects use the LogPad daily to answer questions, which are scored in real time; if the score crosses certain thresholds, patients are instructed to contact the hospital’s pulmonary call center for instruction.
  • 55.
  • 56.
    Clinician Reaction “This technology, which could easily be used with other diseases, has a truly great impact on patients’ quality of life and the disease cost and burden on society,” Dr. Wissam Chatila Temple Lung Center
  • 57.
    Patient Reaction "Withthe (LogPad) you have a daily routine, seven days a week, so you can't miss it. You can't go wrong … This here keeps me out of the hospital as much as possible. That's what I love about it." Edward Goldwire COPD patient
  • 58.
  • 59.
    Discussion 1 Whatfactors determine which studies would benefit from combining physiologic and PRO data collection?
  • 60.
    Discussion 2 Whatbarriers to implementing parallel data capture do you see in your company?
  • 61.
    Discussion 3 Whatstrategies can be used to overcome the barriers to adoption?