MR785 - PD MED Trial- A randomised assessment of the cost-effectiveness of different classes of drugs for Parkinson's Disease

This Data Sharing Agreement permits the University of Birmingham access to HES, Cause of Death, M&P, Cohort Event Notifications and Flagging Current Status reports that were disseminated under a previous iteration of this Agreement, as well as Cancer Registration Data and Civil Registration (Deaths) data going forward, for a study titled the 'PD MED Trial' - a randomised assessment aimed to establish cost effectiveness and which of three classes of drug, as initial treatment, provides the most effective long-term control of symptoms and best quality of life for people with early Parkinson's disease of the cost-effectiveness of different classes of drugs for Parkinson's Disease.

PD MED is a randomised, pragmatic, open label trial that is one of the largest (1,896 UK patients) and longest running trials (20 years) around the world. Its purpose is to compare the long-term effectiveness and cost-effectiveness of four different classes of Parkinson’s disease (PD) medicines, which are currently prescribed to improve Early (newly or less than 6 months diagnosis) and Later (has PD diagnosis plus motor complications unresponsive to Levodopa (LD) dosing changes or timings) patients’ Parkinsonian symptoms. The medications being assessed are levodopa (LD), dopamine agonists (DA), monoamine oxidase type B inhibitors (MAOBI) and catechol-O-methyltransferase inhibitors (COMTI).

Following analyses of its Early participants after 10 years of follow-up, PD MED published in Lancet in 2014 results indicating that LD is more cost-effective in terms of controlling motor symptoms in patients with Early PD. Also, these results subsequently lead to an update of the 2017 NICE guidelines for the treatment of Early PD. These NICE guidelines changes affect how clinicians across the UK treat Early PD. Because LD is more cost-effective than the other classes of medications, this will be economically more beneficial for the NHS. This already is evidenced by NHS Digital’s Prescription Cost Analysis-England 2018: the cost of dopaminergic drugs used to treat PD in 2018 was £91,428,724 whereas the cost in 2008 was £102,762,426

PD MED completed its follow-up period on 31 Dec 2019. The study now wishes to complete the time to event analyses to determine the effectiveness and cost-effectiveness of the four different classes of PD medicines over 20 years and to see whether admission to care homes is delayed, onset of mental incapacity is delayed and to determine if there is prolongation in the years of life of its participants are decreased by using these medicines. The receipt of mortality data and cancer registrations data will allow the researchers to complete the time to event analyses, while also monitoring the safety of long-term administration of the drugs used.

The legal basis for obtaining NHS Digital data is research task in the public interest and the personal data can be lawfully processed under GDPR Articles 6(1)(e) (“processing is necessary for performance of a task in the public interest”) and 9(2)(i) (“processing is necessary for public interest in the area of public health… ensuring high standards of quality and safety of health care and of medicinal products…”). This processing can be deemed to be in the public interest because the work will allow researchers to determine which class of drug provides the most effective control, with the fewest side-effects, for early and later PD. The findings of this study will likely have an impact on the way individuals with PD are cared for.

The cohort, consisting of 1,925 individuals, all have a confirmed diagnosis of Parkinson’s disease. The data requested will be minimised to these individuals.

Patients were eligible for randomisation in the 'Early' arm if:

1. They were previously untreated for PD and therapeutic intervention was considered appropriate. Patients not thought to require dopaminergic treatment at diagnosis were eligible once it was considered that such treatment becomes necessary; or

2. They had previously been treated with dopaminergic medication, but for less than 6 months, and there was now uncertainty as to which class of drug to use. This randomisation may have entailed stopping, or modifying the previous therapy. This would have been left to the discretion of the investigator.

Patients were randomised to the 'Later' arm if:

1. Patients had PD with motor complications unresponsive to changes in levodopa dose and timing

2. Patients in the 'Early' arm can be re-randomised to the 'Later' arm if motor complications develop while on levodopa therapy.

The University of Birmingham is the sole data controller, with sole autonomy for determining how the data will be processed, and will also process the data for the purposes described in this Agreement.

The University of Glasgow will be a data processor on behalf of the University of Birmingham under an appropriate contract for the purpose of undertaking the economic analysis.

The University of Birmingham established a steering committee for the PD MED trial comprised of individuals from external organisations including: University Hospitals Birmingham, Birmingham City Hospital, the University of Oxford, Health Economics Research Centre in Oxford, University of Glasgow, University of Aberdeen, Southern General Hospital, Royal Devon and Exeter Hospital and Addenbrooke’s Hospital. The role of the steering committee is advisory in respect of the whole PD MED clinical trial. The steering committee members have no controllership for the data under this Agreement as they cannot instruct or compel the University of Birmingham to process the data in any way. The steering committee has disbanded but out of courtesy, the University of Birmingham will notify the remaining committee members who haven’t retired or passed away, the end of study results when analyses are completed.

The study has previously received funding from National Institute for Health Research (NIHR) Health Technology Assessment (HTA). While the study does not currently have funding from NIHR HTA the funding will re-commence once the University of Birmingham has received the renewal of data from NHS Digital.

None of the work will take place outside of the UK.

PPIE activities will include informing cohort members news about publications and the trial’s final results through newsletter articles and letters addressed to those cohort members who are still alive. . A large percentage of the PD MED cohort has passed away. By receiving the updated mortality data from NHS Digital, the University of Birmingham will know to whom they can still send the newsletters and letters to update participants. The University of Birmingham plan to send a newsletter out to participants informing them of the publication of the JAMA Neurology paper and their plans for the completion of the trial, once they have received the mortality data from NHS Digital. The public will be notified of the results of the PD MED trial via the PD MED website as well as via press releases.

Results, via email, will also be disseminated to clinicians and research nurses, who will be sent copies of the paper(s) when published. It was initially the intension that a collaborators’ meeting would be held to disseminate the final findings; however, this meeting had to be cancelled due to funding limitations. Final results will thus be emailed to the collaborators once analyses are completed.

Outputs include submission of the write-ups of results to peer-reviewed journals. The National Institute for Clinical Excellence (NICE) will be notified of the results. Results will also be disseminated to the general public via the press. All data will be presented as aggregate data with small numbers suppressed and statistical summaries and measures of association. No patient personal details will be revealed.

These outputs achieve the stated purpose and thus reveal the benefits of processing the data shared by NHS Digital because it is expected to notify clinicians about which class of medicine is most cost-effective to treat patients with Parkinson's disease while improving patient care. By notifying participants, this will inform participants that their participation in the PD MED trial has been essential in determining the improved care of Parkinson's disease patients as well as the fact that clinical trials are useful for improving patient care.

The PD MED cohort consists of Parkinson's Disease patients with 'Early' and 'Later' disease. 'Early' disease patients have newly or recently (< 6 months) diagnosed Parkinson's. 'Later' disease patients already have a diagnosis of Parkinson's but have motor complications from their Parkinson's and are unresponsive to the changes in their levodopa medications. The Later paper has been accepted to JAMA Neurology (entitled, ‘Long term effectiveness of adding COMT or MAOB inhibitors compared to dopamine agonists in poorly controlled Parkinson’s disease-the PDMED trial’) and is expected to be published in early 2022. These results will relate to patients who are in a later stage of Parkinson’s disease. This publication will provide a detailed look at the impact, after 10 years of follow-up, what the four different classes of Parkinson’s medication have on people with Later PD (Parkinson’s with motor complications).

The following papers are currently awaiting submission (all data will be presented as aggregate data) or have been recently published:

i) Hospitalisation following different initial therapies in early Parkinson’s disease: analysis of Hospital Episodes Statistics from the PD MED EARLY trial.

Muzerengi S, Woolley RL, Rick C, Ives N, Dowling F, Gray R, Clarke CE, on behalf of the PD MED Collaborative Group.

Expected publication: Autumn 2022.

ii) Long-term Effectiveness of Adjuvant Treatment With Catechol-O-Methyltransferase or Monoamine Oxidase B Inhibitors

Compared With Dopamine Agonists Among Patients With Parkinson Disease Uncontrolled by Levodopa Therapy The PD MED Randomized Clinical Trial.

Published online on 28 December 2021 in JAMA Neurology, doi:10.1001/jamaneurol.2021.4736.

iii) Cost-effectiveness of dopamine agonists and monoamine oxidase B inhibitors compared with levodopa. Submitted to Movement Disorders but rejected. Currently being resubmitted with rebuttal.

Update: Has been published in the journal 'Movement Disorders', Vol 36, Issue 9, p.2136-2143. 7 May 2021.

The outputs for both PD MED follow-up results, hospitalisation and health economic research will be presented as posters at conferences and published in scientific journals.

Several outputs have already been produced as a result of this research, for example:

1. Mapping from the Parkinson’s Disease Questionnaire PDQ-39 to the Generic EuroQol EQ-5D-3L: The Value of Mixture Models. (April 2015)

2. Comparing results from long and short form versions of the Parkinson's disease questionnaire in a longitudinal study. Jenkinson C, Clarke C, Gray R, Hewitson P, Ives N, Morley D, Rick C, Wheatley K, and others. Parkinsonism & Related Disorders, Vol. 21, Issue 11, p1312–1316 (September 2015)

3. Broadening the evaluative scope of quality of life in Parkinson's: testing the construct validity of the ICECAPO instrument. (June 2016)

4. Too broad to be sensitive? An exploration of the responsiveness of the ICECAP-O capability wellbeing measure compared to the EQ-5D-3L to the change of clinical and QoL aspects in people with Parkinson’s (June 2017).

All outputs will only contain results in highly aggregated format and as statistical summaries and measures of association. Small numbers will be suppressed in line with the HES Analysis Guide. Record level information will not be released to any third party.

The first PD MED results led to a change in the prescribing patterns of the medications for those people with Parkinson’s in the UK.

These results indicate that levodopa is the more effective treatment and more cost-effective in terms of controlling motor complications in patients with Early Parkinson’s disease. Conclusions from the Lancet paper also indicate that after 7 years of follow-up, there are no cumulative adverse effects with LD therapy as well as no loss of benefit with time. Benefits are seen in LD treated patients despite development of more involuntary movements; however, notably, motor fluctuations are not increased.

In addition, these results have led to an update in the 2017 NICE guidelines for the treatment of Early Parkinson’s disease. These NICE guideline changes are expected to affect how clinicians across the UK treat Early Parkinson’s disease – when presented with patients with Early onset Parkinson’s disease, they should begin treatment using levodopa rather than with the other classes of medications. Because levodopa is more cost-effective than the other classes of medications, this should be economically more beneficial for the NHS. After reviewing the Prescription Cost Analysis for England in 2018, it is notable that the net ingredient cost (NIC in sterling pounds) of medicines used to treat Parkinson's disease has been decreasing since the 2017 NICE guideline changes: from £103,428,224 in 2016 to £91,428,724 in 2018.

This Data Sharing Agreement permits the University of Birmingham access to HES, Cause of Death, M&P, Cohort Event Notifications and Flagging Current Status reports that were disseminated under a previous iteration of this Agreement, as well as Cancer Registration Data and Civil Registration (Deaths) data going forward, for a study titled the 'PD MED Trial' - a randomised assessment aimed to establish cost effectiveness and which of three classes of drug, as initial treatment, provides the most effective long-term control of symptoms and best quality of life for people with early Parkinson's disease of the cost-effectiveness of different classes of drugs for Parkinson's Disease.

PD MED is a randomised, pragmatic, open label trial that is one of the largest (1,896 UK patients) and longest running trials (20 years) around the world. Its purpose is to compare the long-term effectiveness and cost-effectiveness of four different classes of Parkinson’s disease (PD) medicines, which are currently prescribed to improve Early (newly or less than 6 months diagnosis) and Later (has PD diagnosis plus motor complications unresponsive to Levodopa (LD) dosing changes or timings) patients’ Parkinsonian symptoms. The medications being assessed are levodopa (LD), dopamine agonists (DA), monoamine oxidase type B inhibitors (MAOBI) and catechol-O-methyltransferase inhibitors (COMTI).

Following analyses of its Early participants after 10 years of follow-up, PD MED published in Lancet in 2014 results indicating that LD is more cost-effective in terms of controlling motor symptoms in patients with Early PD. Also, these results subsequently lead to an update of the 2017 NICE guidelines for the treatment of Early PD. These NICE guidelines changes affect how clinicians across the UK treat Early PD. Because LD is more cost-effective than the other classes of medications, this will be economically more beneficial for the NHS. This already is evidenced by NHS Digital’s Prescription Cost Analysis-England 2018: the cost of dopaminergic drugs used to treat PD in 2018 was £91,428,724 whereas the cost in 2008 was £102,762,426

PD MED completed its follow-up period on 31 Dec 2019. The study now wishes to complete the time to event analyses to determine the effectiveness and cost-effectiveness of the four different classes of PD medicines over 20 years and to see whether admission to care homes is delayed, onset of mental incapacity is delayed and to determine if there is prolongation in the years of life of its participants are decreased by using these medicines. The receipt of mortality data and cancer registrations data will allow the researchers to complete the time to event analyses, while also monitoring the safety of long-term administration of the drugs used.

The legal basis for obtaining NHS Digital data is research task in the public interest and the personal data can be lawfully processed under GDPR Articles 6(1)(e) (“processing is necessary for performance of a task in the public interest”) and 9(2)(i) (“processing is necessary for public interest in the area of public health… ensuring high standards of quality and safety of health care and of medicinal products…”). This processing can be deemed to be in the public interest because the work will allow researchers to determine which class of drug provides the most effective control, with the fewest side-effects, for early and later PD. The findings of this study will likely have an impact on the way individuals with PD are cared for.

The cohort, consisting of 1,925 individuals, all have a confirmed diagnosis of Parkinson’s disease. The data requested will be minimised to these individuals.

Patients were eligible for randomisation in the 'Early' arm if:

1. They were previously untreated for PD and therapeutic intervention was considered appropriate. Patients not thought to require dopaminergic treatment at diagnosis were eligible once it was considered that such treatment becomes necessary; or

2. They had previously been treated with dopaminergic medication, but for less than 6 months, and there was now uncertainty as to which class of drug to use. This randomisation may have entailed stopping, or modifying the previous therapy. This would have been left to the discretion of the investigator.

Patients were randomised to the 'Later' arm if:

1. Patients had PD with motor complications unresponsive to changes in levodopa dose and timing

2. Patients in the 'Early' arm can be re-randomised to the 'Later' arm if motor complications develop while on levodopa therapy.

The University of Birmingham is the sole data controller, with sole autonomy for determining how the data will be processed, and will also process the data for the purposes described in this Agreement.

The University of Glasgow will be a data processor on behalf of the University of Birmingham under an appropriate contract for the purpose of undertaking the economic analysis.

The University of Birmingham established a steering committee for the PD MED trial comprised of individuals from external organisations including: University Hospitals Birmingham, Birmingham City Hospital, the University of Oxford, Health Economics Research Centre in Oxford, University of Glasgow, University of Aberdeen, Southern General Hospital, Royal Devon and Exeter Hospital and Addenbrooke’s Hospital. The role of the steering committee is advisory in respect of the whole PD MED clinical trial. The steering committee members have no controllership for the data under this Agreement as they cannot instruct or compel the University of Birmingham to process the data in any way. The steering committee has disbanded but out of courtesy, the University of Birmingham will notify the remaining committee members who haven’t retired or passed away, the end of study results when analyses are completed.

The study has previously received funding from National Institute for Health Research (NIHR) Health Technology Assessment (HTA). While the study does not currently have funding from NIHR HTA the funding will re-commence once the University of Birmingham has received the renewal of data from NHS Digital.

None of the work will take place outside of the UK.

PPIE activities will include informing cohort members news about publications and the trial’s final results through newsletter articles and letters addressed to those cohort members who are still alive. . A large percentage of the PD MED cohort has passed away. By receiving the updated mortality data from NHS Digital, the University of Birmingham will know to whom they can still send the newsletters and letters to update participants. The University of Birmingham plan to send a newsletter out to participants informing them of the publication of the JAMA Neurology paper and their plans for the completion of the trial, once they have received the mortality data from NHS Digital. The public will be notified of the results of the PD MED trial via the PD MED website as well as via press releases.

Results, via email, will also be disseminated to clinicians and research nurses, who will be sent copies of the paper(s) when published. It was initially the intension that a collaborators’ meeting would be held to disseminate the final findings; however, this meeting had to be cancelled due to funding limitations. Final results will thus be emailed to the collaborators once analyses are completed.

Outputs include submission of the write-ups of results to peer-reviewed journals. The National Institute for Clinical Excellence (NICE) will be notified of the results. Results will also be disseminated to the general public via the press. All data will be presented as aggregate data with small numbers suppressed and statistical summaries and measures of association. No patient personal details will be revealed.

These outputs achieve the stated purpose and thus reveal the benefits of processing the data shared by NHS Digital because it is expected to notify clinicians about which class of medicine is most cost-effective to treat patients with Parkinson's disease while improving patient care. By notifying participants, this will inform participants that their participation in the PD MED trial has been essential in determining the improved care of Parkinson's disease patients as well as the fact that clinical trials are useful for improving patient care.

The PD MED cohort consists of Parkinson's Disease patients with 'Early' and 'Later' disease. 'Early' disease patients have newly or recently (< 6 months) diagnosed Parkinson's. 'Later' disease patients already have a diagnosis of Parkinson's but have motor complications from their Parkinson's and are unresponsive to the changes in their levodopa medications. The Later paper has been accepted to JAMA Neurology (entitled, ‘Long term effectiveness of adding COMT or MAOB inhibitors compared to dopamine agonists in poorly controlled Parkinson’s disease-the PDMED trial’) and is expected to be published in early 2022. These results will relate to patients who are in a later stage of Parkinson’s disease. This publication will provide a detailed look at the impact, after 10 years of follow-up, what the four different classes of Parkinson’s medication have on people with Later PD (Parkinson’s with motor complications).

The following papers are currently awaiting submission (all data will be presented as aggregate data) or have been recently published:

i) Hospitalisation following different initial therapies in early Parkinson’s disease: analysis of Hospital Episodes Statistics from the PD MED EARLY trial.

Muzerengi S, Woolley RL, Rick C, Ives N, Dowling F, Gray R, Clarke CE, on behalf of the PD MED Collaborative Group.

Expected publication: Autumn 2022.

ii) Long-term Effectiveness of Adjuvant Treatment With Catechol-O-Methyltransferase or Monoamine Oxidase B Inhibitors

Compared With Dopamine Agonists Among Patients With Parkinson Disease Uncontrolled by Levodopa Therapy The PD MED Randomized Clinical Trial.

Published online on 28 December 2021 in JAMA Neurology, doi:10.1001/jamaneurol.2021.4736.

iii) Cost-effectiveness of dopamine agonists and monoamine oxidase B inhibitors compared with levodopa. Submitted to Movement Disorders but rejected. Currently being resubmitted with rebuttal.

Update: Has been published in the journal 'Movement Disorders', Vol 36, Issue 9, p.2136-2143. 7 May 2021.

The outputs for both PD MED follow-up results, hospitalisation and health economic research will be presented as posters at conferences and published in scientific journals.

Several outputs have already been produced as a result of this research, for example:

1. Mapping from the Parkinson’s Disease Questionnaire PDQ-39 to the Generic EuroQol EQ-5D-3L: The Value of Mixture Models. (April 2015)

2. Comparing results from long and short form versions of the Parkinson's disease questionnaire in a longitudinal study. Jenkinson C, Clarke C, Gray R, Hewitson P, Ives N, Morley D, Rick C, Wheatley K, and others. Parkinsonism & Related Disorders, Vol. 21, Issue 11, p1312–1316 (September 2015)

3. Broadening the evaluative scope of quality of life in Parkinson's: testing the construct validity of the ICECAPO instrument. (June 2016)

4. Too broad to be sensitive? An exploration of the responsiveness of the ICECAP-O capability wellbeing measure compared to the EQ-5D-3L to the change of clinical and QoL aspects in people with Parkinson’s (June 2017).

All outputs will only contain results in highly aggregated format and as statistical summaries and measures of association. Small numbers will be suppressed in line with the HES Analysis Guide. Record level information will not be released to any third party.

The first PD MED results led to a change in the prescribing patterns of the medications for those people with Parkinson’s in the UK.

These results indicate that levodopa is the more effective treatment and more cost-effective in terms of controlling motor complications in patients with Early Parkinson’s disease. Conclusions from the Lancet paper also indicate that after 7 years of follow-up, there are no cumulative adverse effects with LD therapy as well as no loss of benefit with time. Benefits are seen in LD treated patients despite development of more involuntary movements; however, notably, motor fluctuations are not increased.

In addition, these results have led to an update in the 2017 NICE guidelines for the treatment of Early Parkinson’s disease. These NICE guideline changes are expected to affect how clinicians across the UK treat Early Parkinson’s disease – when presented with patients with Early onset Parkinson’s disease, they should begin treatment using levodopa rather than with the other classes of medications. Because levodopa is more cost-effective than the other classes of medications, this should be economically more beneficial for the NHS. After reviewing the Prescription Cost Analysis for England in 2018, it is notable that the net ingredient cost (NIC in sterling pounds) of medicines used to treat Parkinson's disease has been decreasing since the 2017 NICE guideline changes: from £103,428,224 in 2016 to £91,428,724 in 2018.

This Data Sharing Agreement permits the University of Birmingham continued access to HES, Cause of Death, M&P, Cohort Event Notifications and Flagging Current Status reports that were disseminated under a previous iteration of this Agreement.

PD MED is a randomised, pragmatic, open label trial that is one of the largest (1,896 UK patients) and longest running trials (20 years) around the world. Its purpose is to compare the long-term cost-effectiveness of four different classes of Parkinson’s disease (PD) medicines, which are currently prescribed to improve Early (newly or less than 6 months diagnosis) and Later (has PD diagnosis plus motor complications unresponsive to LD dosing changes or timings) patients’ Parkinsonian symptoms. The medications being assessed are levodopa (LD), dopamine agonists (DA), monoamine oxidase type B inhibitors (MAOBI) and catechol-O-methyltransferase inhibitors (COMTI).

Following analyses of its Early participants after 10 years of follow-up, PD MED published in Lancet in 2014 results indicating that LD is more cost-effective in terms of controlling motor symptoms in patients with Early PD. Also, these results subsequently lead to an update of the 2017 NICE guidelines for the treatment of Early PD. These NICE guidelines changes affect how clinicians across the UK treat Early PD. Because LD is more cost-effective than the other classes of medications, this will be economically more beneficial for the NHS. This already is evidenced by NHS Digital’s Prescription Cost Analysis-England 2018: the cost of dopaminergic drugs used to treat PD in 2018 was £91,428,724 whereas the cost in 2008 was £102,762,426

PD MED completed its follow-up period on 31 Dec 2019. The study now wishes to complete the time to event analyses to determine the cost-effectiveness of the four different classes of PD medicines over 20 years and to see whether institutionalisation, onset of dementia and death of its participants are decreased by using these medicines. The receipt of death data, cancer registrations and patient demographics will allow the researchers complete the time to event analyses, while also monitoring the safety of long-term administration of the drugs used.

The legal basis for obtaining NHS Digital data is research task in the public interest and the personal data can be lawfully processed under GDPR Articles 6(1)(e) (“processing is necessary for performance of a task in the public interest”) and 9(2)(i) (“processing is necessary for public interest in the area of public health… ensuring high standards of quality and safety of health care and of medicinal products…”). This processing can be deemed to be in the public interest because the work will allow researchers to determine which class of drug provides the most effective control, with the fewest side-effects, for early and later PD. The findings of this study will likely have an impact on the way individuals with PD are cared for.

The cohort, consisting of 1,925 individuals, all have a confirmed diagnosis of Parkinson’s disease. The data requested will be minimised to these individuals.

Patients were eligible for randomisation in the 'Early' arm if:

1. They were previously untreated for PD and therapeutic intervention was considered appropriate. Patients not thought to require dopaminergic treatment at diagnosis were eligible once it was considered that such treatment becomes necessary; or

2. They had previously been treated with dopaminergic medication, but for less than 6 months, and there was now uncertainty as to which class of drug to use. This randomisation may have entailed stopping, or modifying the previous therapy. This would have been left to the discretion of the investigator.

Patients were randomised to the 'Later' arm if:

1. Patients had PD with motor complications unresponsive to changes in levodopa dose and timing

2. Patients in the 'Early' arm can be re-randomised to the 'Later' arm if motor complications develop while on levodopa therapy.

The University of Birmingham is the sole data controller, with sole autonomy for determining how the data will be processed, and will also process the data for the purposes described in this agreement.

The University of Glasgow will be a data processor on behalf of the University of Birmingham under an appropriate contract for the purpose of undertaking the economic analysis.

The University of Birmingham established a steering committee for the PD MED trial comprised of individuals from external organisations including: University Hospitals Birmingham, Birmingham City Hospital, Radcliffe Infirmary in Oxford, the University of Oxford, Health Economics Research Centre in Oxford, North Tyneside General Hospital, St Anna’s Hospital in Brno in the Czech Republic, Russian State Medical University in Moscow in Russia, and Grampian University Hospital in Aberdeen. The role of the steering committee is advisory in respect of the whole PD MED clinical trial. The steering committee members have no controllership for the data under this Agreement as they cannot instruct or compel the University of Birmingham to process the data in any way.

The study has previously received funding from NIHR HTA. While the study does not currently have funding from NIHR HTA the funding will re-commence once the University of Birmingham has received the renewal of data from NHS Digital.

None of the work will take place outside of the UK.

Results from the PD MED study will be disseminated to the participants of the PD MED clinical trial through newsletter articles and letters addressed to those cohort members who are still alive. The public will be notified of the results of the PD MED trial via the PD MED website. Results will also be disseminated to clinicians and research nurses via email and a collaborator's meeting, which will be organised for spring 2021. Speakers at the collaborator's meeting will explain the results from the ‘Later’ paper as well as on results from the analyses of Early and Later disease data after 20 years of follow-up.

Outputs include submission of the write-ups of results to peer-reviewed journals and should the results be impactful, the results will be submitted to the National Institute for Clinical Excellence (NICE). All data will be presented as aggregate data and statistical summaries and measures of association. No patient personal details will be revealed.

These outputs achieve the stated purpose and thus reveal the benefits of processing the data shared by NHS Digital because it will notify clinicians about which class of medicine is most cost-effective to treat patients with Parkinson's disease while improving patient care. By notifying participants, this will inform participants that their participation in the PD MED trial has been essential in determining the improved care of Parkinson's disease patients as well as the fact that clinical trials are useful for improving patient care.

The PD MED cohort consists of Parkinson's Disease patients with 'Early' and 'Later' disease. 'Early' disease patients have newly or recently (< 6 months) diagnosed Parkinson's. 'Later' disease patients already have a diagnosis of Parkinson's but have motor complications from their Parkinson's and are unresponsive to the changes in their levodopa medications. The target date for publication of the first Later disease results is autumn 2020 (These results will relate to patients who are in a later stage of the disease). This publication will provide a detailed look at the impact, after 10 years of follow-up, that the four different classes of Parkinson’s medication have on people with Later PD (Parkinson’s with motor complications). This paper was submitted to a journal but was rejected. It has since been reconfigured for JAMA Neurology and will imminently be submitted for publication.

Should any results be impactful and newsworthy, the study team will also notify the press so that the general public will be aware of the results

The following papers are currently awaiting submission (all data will be presented as aggregate data):

i) Hospitalisation following different initial therapies in early Parkinson’s disease: analysis of Hospital Episodes Statistics from the PD MED EARLY trial.

Muzerengi S, Woolley RL, Rick C, Ives N, Dowling F, Gray R, Clarke CE, on behalf of the PD MED Collaborative Group.

Expected publication: by winter 2020

ii) Dopamine agonist versus monoamine oxidase B inhibitor or catechol-O-methyl transferase inhibitor therapy in later Parkinson's disease: a large pragmatic randomised controlled trial (PD MED LATER). PD MED Collaborative Group.

A draft paper was rejected from an initial journal but researchers have now reformatted it for JAMA Neurology and it will be submitted soon.

iii) Cost-effectiveness of dopamine agonists and monoamine oxidase B inhibitors compared with levodopa. Submitted to Movement Disorders but rejected. Currently being resubmitted with rebuttal.

The outputs for both PD MED follow-up results, hospitalisation and health economic research will be presented as posters at conferences and published in scientific journals.

Several outputs have already been produced as a result of this research:

1. Evaluating Drug Treatments for Parkinson's Disease. BMJ. 2002 June; 324:1508-11. Wheatley, K., Stowe, R., Clarke, C.E., Gray, R., et. al.

2. Systemic Review of Levodopa Dose Equivalency Reporting in Parkinson's Disease. Movement Disorder. 2010 Nov; 25(15):2649-53. Tomlinson, C., Harrison, R., Gray, R., Rick, C., et. al.

3. Long-term effectiveness of dopamine agonists and monoamine oxidase B inhibitors compared to levodopa as initial therapy of Parkinson's disease: PD MED, a large pragmatic randomised trial. PD MED Collaborative Group (June 2014)

https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(14)60683-8/fulltext

4. Mapping from the Parkinson’s Disease Questionnaire PDQ-39 to the Generic EuroQol EQ-5D-3L: The Value of Mixture Models. (April 2015)

5. Comparing results from long and short form versions of the Parkinson's disease questionnaire in a longitudinal study. Jenkinson C, Clarke C, Gray R, Hewitson P, Ives N, Morley D, Rick C, Wheatley K, and others. Parkinsonism & Related Disorders, Vol. 21, Issue 11, p1312–1316 (September 2015)

6. Broadening the evaluative scope of quality of life in Parkinson's: testing the construct validity of the ICECAPO instrument. (June 2016)

7. Too broad to be sensitive? An exploration of the responsiveness of the ICECAP-O capability wellbeing measure compared to the EQ-5D-3L to the change of clinical and QoL aspects in people with Parkinson’s (June 2017).

All outputs will only contain results in highly aggregated format and as statistical summaries and measures of association. Small numbers will be suppressed in line with the HES Analysis Guide. Record level information will not be released to any third party.

The first PD MED results led to a change in the prescribing patterns of the medications for those people with Parkinson’s in the UK.

These results indicate that levodopa is more cost-effective in terms of controlling motor complications in patients with Early Parkinson’s disease. In addition, these results have led to an update in the 2017 NICE guidelines for the treatment of Early Parkinson’s disease. These NICE guideline changes affect how clinicians across the UK will treat Early Parkinson’s disease – when presented with patients with Early onset Parkinson’s disease, they will begin treatment using levodopa rather than with the other classes of medications. Because levodopa is more cost-effective than the other classes of medications, this will be economically more beneficial for the NHS. After reviewing the Prescription Cost Analysis for England in 2018, it is notable that the net ingredient cost (NIC in sterling pounds) of medicines used to treat Parkinson's disease has been decreasing since the 2017 NICE guideline changes: from £103,428,224 in 2016 to £91,428,724 in 2018.

This Data Sharing Agreement permits the retention of the data provided under previous iterations of this Agreement for an interim period. This is a pragmatic approach to provide an active Agreement whilst enabling University of Birmingham to complete the necessary actions to enable a subsequent application to extend the Agreement meeting all applicable data sharing standards as published in NHS Digital’s website (see: https://digital.nhs.uk/services/data-access-request-service-dars/dars-guidance).

The following provides background information on the purpose of the original study:

The University of Birmingham request continuing access to HES, Cause of Death, M&P, Cohort Event Notifications and Flagging Current Status reports that were disseminated under a previous iteration of this Agreement.

PD MED is a randomised, pragmatic, open label trial that is one of the largest (1,896 UK patients) and longest running trials (20 years) around the world. Its purpose is to compare the long-term cost-effectiveness of four different classes of Parkinson’s disease (PD) medicines, which are currently prescribed to improve Early (newly or less than 6 months diagnosis) and Later (has PD diagnosis plus motor complications unresponsive to LD dosing changes or timings) patients’ Parkinsonian symptoms. The medications being assessed are levodopa (LD), dopamine agonists (DA), monoamine oxidase type B inhibitors (MAOBI) and catechol-O-methyltransferase inhibitors (COMTI).

Following analyses of its Early participants after 10 years of follow-up, PD MED published in Lancet in 2014 results indicating that LD is more cost-effective in terms of controlling motor symptoms in patients with Early PD. Also, these results subsequently lead to an update of the 2017 NICE guidelines for the treatment of Early PD. These NICE guidelines changes affect how clinicians across the UK treat Early PD. Because LD is more cost-effective than the other classes of medications, this will be economically more beneficial for the NHS. This already is evidenced by NHS Digital’s Prescription Cost Analysis-England 2018: the cost of dopaminergic drugs used to treat PD in 2018 was £91,428,724 whereas the cost in 2008 was £102,762,426

PD MED completed its follow-up period on 31 Dec 2019. The study now wishes to complete the time to event analyses to determine the cost-effectiveness of the four different classes of PD medicines over 20 years and to see whether institutionalisation, onset of dementia and death of its participants are decreased by using these medicines. The receipt of death data, cancer registrations and patient demographics will allow the researchers complete the time to event analyses, while also monitoring the safety of long-term administration of the drugs used.

The legal basis for obtaining NHS Digital data is research task in the public interest and the personal data can be lawfully processed under GDPR Articles 6(1)(e) (“processing is necessary for performance of a task in the public interest”) and 9(2)(i) (“processing is necessary for public interest in the area of public health… ensuring high standards of quality and safety of health care and of medicinal products…”). This processing can be deemed to be in the public interest because the work will allow researchers to determine which class of drug provides the most effective control, with the fewest side-effects, for early and later PD. The findings of this study will likely have an impact on the way individuals with PD are cared for.

The cohort, consisting of 1896 individuals, all have a confirmed diagnosis of Parkinson’s disease. The data requested will be minimised to these individuals.

Patients were eligible for randomisation in the 'Early' arm if:

1. They were previously untreated for PD and therapeutic intervention was considered appropriate. Patients not thought to require dopaminergic treatment at diagnosis were eligible once it was considered that such treatment becomes necessary; or

2. They had previously been treated with dopaminergic medication, but for less than 6 months, and there was now uncertainty as to which class of drug to use. This randomisation may have entailed stopping, or modifying the previous therapy. This would have been left to the discretion of the investigator.

Patients were randomised to the 'Later' arm if:

1. Patients had PD with motor complications unresponsive to changes in levodopa dose and timing

2. Patients in the 'Early' arm can be re-randomised to the 'Later' arm if motor complications develop while on levodopa therapy.

The University of Birmingham is the sole data controller, with sole autonomy for determining how the data will be processed, and will also process the data for the purposes described in this agreement.

The University of Glasgow will be a data processor on behalf of the University of Birmingham under an appropriate contract for the purpose of undertaking the economic analysis.

The University of Birmingham established a steering committee for the PD MED trial comprised of individuals from external organisations including: University Hospitals Birmingham, Birmingham City Hospital, Radcliffe Infirmary in Oxford, the University of Oxford, Health Economics Research Centre in Oxford, North Tyneside General Hospital, St Anna’s Hospital in Brno in the Czech Republic, Russian State Medical University in Moscow in Russia, and Grampian University Hospital in Aberdeen. The role of the steering committee is advisory in respect of the whole PD MED clinical trial. The steering committee members have no controllership for the data under this Agreement as they cannot instruct or compel the University of Birmingham to process the data in any way.

The study has previously received funding from NIHR HTA. While the study does not currently have funding from NIHR HTA the funding will re-commence once the University of Birmingham has received the renewal of data from NHS Digital.

None of the work will take place outside of the UK.

This Agreement permits the secure retention of the data only and no other processing.

No new outputs will be produced under this Data Sharing Agreement.

Results from the PD MED study will be disseminated to the participants of the PD MED clinical trial through newsletter articles and letters addressed to those cohort members who are still alive. The public will be notified of the results of the PD MED trial via the PD MED website. Results will also be disseminated to clinicians and research nurses via email and a collaborator's meeting, which will be organised for spring 2021. Speakers at the collaborator's meeting will explain the results from the ‘Later’ paper as well as on results from the analyses of Early and Later disease data after 20 years of follow-up.

Outputs include submission of the write-ups of results to peer-reviewed journals and should the results be impactful, the results will be submitted to the National Institute for Clinical Excellence (NICE). All data will be presented as aggregate data and statistical summaries and measures of association. No patient personal details will be revealed.

These outputs achieve the stated purpose and thus reveal the benefits of processing the data shared by NHS Digital because it will notify clinicians about which class of medicine is most cost-effective to treat patients with Parkinson's disease while improving patient care. By notifying participants, this will inform participants that their participation in the PD MED trial has been essential in determining the improved care of Parkinson's disease patients as well as the fact that clinical trials are useful for improving patient care.

The PD MED cohort consists of Parkinson's Disease patients with 'Early' and 'Later' disease. 'Early' disease patients have newly or recently (< 6 months) diagnosed Parkinson's. 'Later' disease patients already have a diagnosis of Parkinson's but have motor complications from their Parkinson's and are unresponsive to the changes in their levodopa medications. The target date for publication of the first Later disease results is autumn 2020 (These results will relate to patients who are in a later stage of the disease). This publication will provide a detailed look at the impact, after 10 years of follow-up, that the four different classes of Parkinson’s medication have on people with Later PD (Parkinson’s with motor complications). This paper was submitted to a journal but was rejected. It has since been reconfigured for JAMA Neurology and will imminently be submitted for publication.

Should any results be impactful and newsworthy, we will also notify the press so that the general public will be aware of our results

The following papers are currently awaiting submission (all data will be presented as aggregate data):

i) Hospitalisation following different initial therapies in early Parkinson’s disease: analysis of Hospital Episodes Statistics from the PD MED EARLY trial.

Muzerengi S, Woolley RL, Rick C, Ives N, Dowling F, Gray R, Clarke CE, on behalf of the PD MED Collaborative Group.

Expected publication: by winter 2020

ii) Dopamine agonist versus monoamine oxidase B inhibitor or catechol-O-methyl transferase inhibitor therapy in later Parkinson's disease: a large pragmatic randomised controlled trial (PD MED LATER). PD MED Collaborative Group.

A draft paper was rejected from an initial journal but researchers have now reformatted it for JAMA Neurology and it will be submitted soon.

iii) Cost-effectiveness of dopamine agonists and monoamine oxidase B inhibitors compared with levodopa. Submitted to Movement Disorders but rejected. Currently being resubmitted with rebuttal.

The outputs for both PD MED follow-up results, hospitalisation and health economic research will be presented as posters at conferences and published in scientific journals.

Several outputs have already been produced as a result of this research:

1. Evaluating Drug Treatments for Parkinson's Disease. BMJ. 2002 June; 324:1508-11. Wheatley, K., Stowe, R., Clarke, C.E., Gray, R., et. al.

2. Systemic Review of Levodopa Dose Equivalency Reporting in Parkinson's Disease. Movement Disorder. 2010 Nov; 25(15):2649-53. Tomlinson, C., Harrison, R., Gray, R., Rick, C., et. al.

3. Long-term effectiveness of dopamine agonists and monoamine oxidase B inhibitors compared to levodopa as initial therapy of Parkinson's disease: PD MED, a large pragmatic randomised trial. PD MED Collaborative Group (June 2014)

https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(14)60683-8/fulltext

4. Mapping from the Parkinson’s Disease Questionnaire PDQ-39 to the Generic EuroQol EQ-5D-3L: The Value of Mixture Models. (April 2015)

5. Comparing results from long and short form versions of the Parkinson's disease questionnaire in a longitudinal study. Jenkinson C, Clarke C, Gray R, Hewitson P, Ives N, Morley D, Rick C, Wheatley K, and others. Parkinsonism & Related Disorders, Vol. 21, Issue 11, p1312–1316 (September 2015)

6. Broadening the evaluative scope of quality of life in Parkinson's: testing the construct validity of the ICECAPO instrument. (June 2016)

7. Too broad to be sensitive? An exploration of the responsiveness of the ICECAP-O capability wellbeing measure compared to the EQ-5D-3L to the change of clinical and QoL aspects in people with Parkinson’s (June 2017).

All outputs will only contain results in highly aggregated format and as statistical summaries and measures of association. Small numbers will be suppressed in line with the HES Analysis Guide. Record level information will not be released to any third party.

The first PD MED results led to a change in the prescribing patterns of the medications for those people with Parkinson’s in the UK.

These results indicate that levodopa is more cost-effective in terms of controlling motor complications in patients with Early Parkinson’s disease. In addition, these results have led to an update in the 2017 NICE guidelines for the treatment of Early Parkinson’s disease. These NICE guideline changes affect how clinicians across the UK will treat Early Parkinson’s disease – when presented with patients with Early onset Parkinson’s disease, they will begin treatment using levodopa rather than with the other classes of medications. Because levodopa is more cost-effective than the other classes of medications, this will be economically more beneficial for the NHS. After reviewing the Prescription Cost Analysis for England in 2018, it is notable that the net ingredient cost (NIC in sterling pounds) of medicines used to treat Parkinson's disease has been decreasing since the 2017 NICE guideline changes: from £103,428,224 in 2016 to £91,428,724 in 2018.

This Data Sharing Agreement permits the retention of the data provided under previous iterations of this Agreement for an interim period. This is a pragmatic approach to provide an active Agreement whilst enabling University of Birmingham to complete the necessary actions to enable a subsequent application to extend the Agreement meeting all applicable data sharing standards as published in NHS Digital’s website (see: https://digital.nhs.uk/services/data-access-request-service-dars/dars-guidance).

The following provides background information on the purpose of the original study:

The University of Birmingham requires continuing access to HES data and mortality, patient demographics and cancer registration data for the purpose of the PD MED clinical trial. This trial started in 2000 and is a large randomised controlled trial comparing the effectiveness and cost-effectiveness of four classes of medication for people with Parkinson’s disease.

The trial design collected limited hospitalisation data, with targeted serious adverse event (SAE) reporting. Only SAE’s that were thought to be related to medication were requested, as, given the population, (average age at randomisation was 70) many non-treatment related hospitalisations would occur.

The University of Birmingham will examine hospitalisations and deaths following different initial therapies to treat Parkinson's Disease. The data may show that one arm (i.e. which PD drug was given) of the trial may have lower or higher amount of hospitalisations. This in turn, if it showed either a benefit or negative impact, would be published.

Giving NHS commissioners this knowledge will hopefully impact the prescribing patterns of UK clinicians as they will be informed of the benefits which in turn will improve the treatment of patients. This has already been demonstrated by the first PD MED results which have changed the prescribing patterns of the medications for those people with Parkinson’s in the UK. Adding hospitalisation data to this would help improve the current guidance.

This is a research task in the public interest and the personal data can be lawfully processed under GDPR Articles 6(1)(e) (“processing is necessary for the performance of a task in the public interest”) and 9(2)(j) (“processing is necessary for… scientific or historical research purposes”).

All participants consented to be in the clinical trial. The trial has ethical approval. Section 251 support was granted to permit access to hospital episodes statistics (HES) and any HES data to be shared with the University of Glasgow under this Agreement will be pseudonymised and appropriately minimised to protect confidentiality.

Mortality, patient demographic and cancer registration data were first collected for the PD MED follow-up study in 2008, with HES being requested in 2013. HES data enabled the collection of detailed hospitalisation data without unreasonable burdens on the clinical staff at hospitals involved in clinical trials. The University of Birmingham therefore requested the hospitalisation and hospital usage data for the PD MED cohort to provide detailed information on all hospitalisations and hospital usage to benefit the results obtainable from the clinical trial. Analysis on this HES data is not yet complete and therefore University of Birmingham wishes to retain the HES data and in the future will request a final report of HES data for the years since 2012/13.

The data subjects are all individuals with confirmed diagnosis of Parkinson’s disease.

Patients were eligible for early randomisation if:

1. They were previously untreated for PD and therapeutic intervention was considered appropriate. Patients not thought to require dopaminergic treatment at diagnosis were eligible once it was considered that such treatment becomes necessary; or

2. They had previously been treated with dopaminergic medication, but for less than 6 months, and there was now uncertainty as to which class of drug to use. This randomisation may have entailed stopping, or modifying the previous therapy. This would have been left to the discretion of the investigator.

Patients who developed motor complications that were uncontrolled by Levodopa (LD) alone or in combination with either dopamine agonists (DA) or monoamine oxidase type B inhibitors (MAOBI) and hence required the addition of another class of drug were eligible for the later disease randomisation.

Consent was obtained from the patient or carer (according to the appropriate practice). Participants were recruited between 1999 and 2009.

The details of the cohort of 1,896 participants are already held by NHS Digital and will not be added to. The trial is due to close in November 2019 and around this time, the University of Birmingham will require further reports of participants’ hospitalisations and/or deaths in order to update the trial results. This data will be requested in a future application to renew this Data Sharing Agreement. No data will be supplied under this Data Sharing Agreement.

The University of Birmingham has received mortality, patient demographic and cancer registration data and HES data under previous Agreements. Under this Agreement, the mortality, patient demographic and cancer registration data will be stored but no new analyses would be undertaken using this data. Further analyses are planned in the future when new mortality data is obtained, subject to a future application demonstrating a clear legal basis to access mortality data. The mortality, patient demographic and cancer registration data supplied periodically until September 2018 has been processed and analysed periodically.

The trial data includes details of patients' treatments during the trial and information collected through postal questionnaires completed by patients and their carers. A subset of trial data including pseudonymised HES data will be securely transferred to the University of Glasgow for the collaborators there to perform a health economic evaluation. The HES data will be used in conjunction with PD MED data set to help better answer the questions of the PD MED study for both clinical and cost effectiveness. This Data Sharing Agreement does not permit any mortality, patient demographic or cancer registration data to be shared with the University of Glasgow.

The University of Birmingham is the sole data controller with sole autonomy for determining that the data under this Agreement is required for the purpose of the PD MED trial and for determining the purposes for and manner of the processing of that data. The University of Birmingham will also process the data.

The University of Glasgow will be a data processor on behalf of the University of Birmingham under an appropriate contract for the purpose of undertaking the economic analysis.

The University of Birmingham established a steering committee for the PD MED trial comprised of individuals from external organisations including: University Hospitals Birmingham, Birmingham City Hospital, Radcliffe Infirmary in Oxford, the University of Oxford, Health Economics Research Centre in Oxford, North Tyneside General Hospital, St Anna’s Hospital in Brno in the Czech Republic, Russian State Medical University in Moscow in Russia, and Grampian University Hospital in Aberdeen. The role of the steering committee is advisory in respect of the whole PD MED clinical trial. The steering committee members have no controllership for the data under this Agreement as they cannot instruct or compel the University of Birmingham to process the data in any way.

The study is an NIHR HTA funded project.

None of the work will take place outside of the UK.

This Agreement permits the secure retention of the data only and no other processing.

No new outputs will be produced under this Data Sharing Agreement.

Giving NHS commissioners this knowledge will impact the prescribing patterns of UK clinicians as they will be informed of the benefits which in turn will improve treatment of patients. This has already been demonstrated by the first PD MED results which have changed the prescribing patterns of the medications for those people with Parkinson’s in the UK

The PD MED trial early results have already changed current clinical practice in the UK, as they were part of the 2017 update to the NICE guidelines for PD. (https://www.nice.org.uk/guidance/ng71/evidence/full-guideline-pdf-4538466253).

Previous publications:

1. Evaluating Drug Treatments for Parkinson's Disease. BMJ. 2002 June; 324:1508-11. Wheatley, K., Stowe, R., Clarke, C.E., Gray, R., et. al.

2. Systemic Review of Levodopa Dose Equivalency Reporting in Parkinson's Disease. Movement Disorder. 2010 Nov; 25(15):2649-53. Tomlinson, C., Harrison, R., Gray, R., Rick, C., et. al.

3. Long-term effectiveness of dopamine agonists and monoamine oxidase B inhibitors compared to levodopa as initial therapy of Parkinson's disease: PD MED, a large pragmatic randomised trial. PD MED Collaborative Group (June 2014)

https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(14)60683-8/fulltext

4. Mapping from the Parkinson’s Disease Questionnaire PDQ-39 to the Generic EuroQol EQ-5D-3L: The Value of Mixture Models. (April 2015)

5. Comparing results from long and short form versions of the Parkinson's disease questionnaire in a longitudinal study. Jenkinson C, Clarke C, Gray R, Hewitson P, Ives N, Morley D, Rick C, Wheatley K, and others. Parkinsonism & Related Disorders, Vol. 21, Issue 11, p1312–1316 (September 2015)

6. Broadening the evaluative scope of quality of life in Parkinson's: testing the construct validity of the ICECAPO instrument. (June 2016)

7. Too broad to be sensitive? An exploration of the responsiveness of the ICECAP-O capability wellbeing measure compared to the EQ-5D-3L to the change of clinical and QoL aspects in people with Parkinson’s (June 2017).

The University of Birmingham requires continuing access to HES data and mortality, patient demographics and cancer registration data for the purpose of the PD MED clinical trial. This trial started in 2000 and is a large randomised controlled trial comparing the effectiveness and cost-effectiveness of four classes of medication for people with Parkinson’s disease.

The trial design collected limited hospitalisation data, with targeted serious adverse event (SAE) reporting. Only SAE’s that were thought to be related to medication were requested, as, given the population, (average age at randomisation was 70) many non-treatment related hospitalisations would occur.

The University of Birmingham will examine hospitalisations and deaths following different initial therapies to treat Parkinson's Disease. The data may show that one arm (i.e. which PD drug was given) of the trial may have lower or higher amount of hospitalisations. This in turn, if it showed either a benefit or negative impact, would be published.

Giving NHS commissioners this knowledge will hopefully impact the prescribing patterns of UK clinicians as they will be informed of the benefits which in turn will improve the treatment of patients. This has already been demonstrated by the first PD MED results which have changed the prescribing patterns of the medications for those people with Parkinson’s in the UK. Adding hospitalisation data to this would help improve the current guidance.

This is a research task in the public interest and the personal data can be lawfully processed under GDPR Articles 6(1)(e) (“processing is necessary for the performance of a task in the public interest”) and 9(2)(j) (“processing is necessary for… scientific or historical research purposes”).

All participants consented to be in the clinical trial. The trial has ethical approval. Section 251 support was granted to permit access to hospital episodes statistics (HES) and any HES data to be shared with the University of Glasgow under this Agreement will be pseudonymised and appropriately minimised to protect confidentiality.

Mortality, patient demographic and cancer registration data were first collected for the PD MED follow-up study in 2008, with HES being requested in 2013. HES data enabled the collection of detailed hospitalisation data without unreasonable burdens on the clinical staff at hospitals involved in clinical trials. The University of Birmingham therefore requested the hospitalisation and hospital usage data for the PD MED cohort to provide detailed information on all hospitalisations and hospital usage to benefit the results obtainable from the clinical trial. Analysis on this HES data is not yet complete and therefore University of Birmingham wishes to retain the HES data and in the future will request a final report of HES data for the years since 2012/13.

The data subjects are all individuals with confirmed diagnosis of Parkinson’s disease.

Patients were eligible for early randomisation if:

1. They were previously untreated for PD and therapeutic intervention was considered appropriate. Patients not thought to require dopaminergic treatment at diagnosis were eligible once it was considered that such treatment becomes necessary; or

2. They had previously been treated with dopaminergic medication, but for less than 6 months, and there was now uncertainty as to which class of drug to use. This randomisation may have entailed stopping, or modifying the previous therapy. This would have been left to the discretion of the investigator.

Patients who developed motor complications that were uncontrolled by Levodopa (LD) alone or in combination with either dopamine agonists (DA) or monoamine oxidase type B inhibitors (MAOBI) and hence required the addition of another class of drug were eligible for the later disease randomisation.

Consent was obtained from the patient or carer (according to the appropriate practice). Participants were recruited between 1999 and 2009.

The details of the cohort of 1,896 participants are already held by NHS Digital and will not be added to. The trial is due to close in November 2019 and around this time, the University of Birmingham will require further reports of participants’ hospitalisations and/or deaths in order to update the trial results. This data will be requested in a future application to renew this Data Sharing Agreement. No data will be supplied under this Data Sharing Agreement.

The University of Birmingham has received mortality, patient demographic and cancer registration data and HES data under previous Agreements. Under this Agreement, the mortality, patient demographic and cancer registration data will be stored but no new analyses would be undertaken using this data. Further analyses are planned in the future when new mortality data is obtained, subject to a future application demonstrating a clear legal basis to access mortality data. The mortality, patient demographic and cancer registration data supplied periodically until September 2018 has been processed and analysed periodically.

The trial data includes details of patients' treatments during the trial and information collected through postal questionnaires completed by patients and their carers. A subset of trial data including pseudonymised HES data will be securely transferred to the University of Glasgow for the collaborators there to perform a health economic evaluation. The HES data will be used in conjunction with PD MED data set to help better answer the questions of the PD MED study for both clinical and cost effectiveness. This Data Sharing Agreement does not permit any mortality, patient demographic or cancer registration data to be shared with the University of Glasgow.

The University of Birmingham is the sole data controller with sole autonomy for determining that the data under this Agreement is required for the purpose of the PD MED trial and for determining the purposes for and manner of the processing of that data. The University of Birmingham will also process the data.

The University of Glasgow will be a data processor on behalf of the University of Birmingham under an appropriate contract for the purpose of undertaking the economic analysis.

The University of Birmingham established a steering committee for the PD MED trial comprised of individuals from external organisations including: University Hospitals Birmingham, Birmingham City Hospital, Radcliffe Infirmary in Oxford, the University of Oxford, Health Economics Research Centre in Oxford, North Tyneside General Hospital, St Anna’s Hospital in Brno in the Czech Republic, Russian State Medical University in Moscow in Russia, and Grampian University Hospital in Aberdeen. The role of the steering committee is advisory in respect of the whole PD MED clinical trial. The steering committee members have no controllership for the data under this Agreement as they cannot instruct or compel the University of Birmingham to process the data in any way.

The study is an NIHR HTA funded project.

None of the work will take place outside of the UK.

The data will be included in analyses of hospitalisations and hospital usage that will be published in peer reviewed journals. Individuals will not be identifiable within these publications.

These results will also be disseminated to the participants in the PD MED clinical trial through newsletter articles.

The target date for publication of the first later disease (motor complications) results are due in autumn 2019. This publication will demonstrate a detailed long term look at the impact that different classes of Parkinson’s medication have on People with Parkinson’s, including the impact the different classes of Parkinson’s medication have on participant hospitalisations.

The following papers are currently being reviewed (all data will be presented as aggregate data):

i) Hospitalisation following different initial therapies in early Parkinson’s disease: analysis of Hospital Episodes Statistics from the PD MED EARLY trial.

Muzerengi S, Woolley RL, Rick C, Ives N, Dowling F, Gray R, Clarke CE, on behalf of the PD MED Collaborative Group.

Expected publication: by autumn 2019.

ii) Dopamine agonist versus monoamine oxidase B inhibitor or catechol-O-methyl transferase inhibitor therapy in later Parkinson's disease: a large pragmatic randomised controlled trial (PD MED LATER). PD MED Collaborative Group.

A draft paper was rejected from initial journal and we are now trying to find somewhere else to submit it spring 2020

The outputs for both the hospitalisation and health economic research will be presented as posters at conferences and published in scientific journals.

All outputs will only contain results in highly aggregated format and as statistical summaries and measures of association. Small numbers will be suppressed in line with the HES Analysis Guide. Record level information will not be released to any third party.

Giving NHS commissioners this knowledge will impact the prescribing patterns of UK clinicians as they will be informed of the benefits which in turn with treatment of patients. This has already been demonstrated by the first PD MED results which have changed the prescribing patterns of the medications for those people with Parkinson’s in the UK

The PD MED trial early results have already changed current clinical practice in the UK, as they were part of the 2017 update to the NICE guidelines for PD. https://www.nice.org.uk/guidance/ng71/evidence/full-guideline-pdf-4538466253).

Previous publications:

Long-term effectiveness of dopamine agonists and monoamine oxidase B inhibitors compared to levodopa as initial therapy of Parkinson's disease: PD MED, a large pragmatic randomised trial. PD MED Collaborative Group (June 2014)

https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(14)60683-8/fulltext

Mapping from the Parkinson’s Disease Questionnaire PDQ-39 to the Generic EuroQol EQ-5D-3L: The Value of Mixture Models. (April 2015)

Comparing results from long and short form versions of the Parkinson's disease questionnaire in a longitudinal study. Jenkinson C, Clarke C, Gray R, Hewitson P, Ives N, Morley D, Rick C, Wheatley K, and others. Parkinsonism & Related Disorders, Vol. 21, Issue 11, p1312–1316 (September 2015)

Randomised trial comparing dopamine agonist, MAOB inhibitor and COMT inhibitor as adjuvant therapy in later Parkinson's disease (PD) (June 2016)

Broadening the evaluative scope of quality of life in Parkinson's: testing the construct validity of the ICECAPO instrument. (June 2016)

Too broad to be sensitive? An exploration of the responsiveness of the ICECAP-O capability wellbeing measure compared to the EQ-5D-3L to the change of clinical and QoL aspects in people with Parkinson’s (June 2017).