OptiMine: Optimizing the mining of electronic health record data to increase uptake of cancer prevention digital interventions

Project Outline/Introduction

Smoking and drinking alcohol at risky levels are associated with cancer later on in life.  In England, 15% of adults (17% men and 13% women) smoke cigarettes, and 24% of adults (31% men and 16% women) drink alcohol at risky levels (>14 units per week.  There are services to help people quit smoking and reduce their alcohol consumption, but barriers to accessing these include time to travel and sensitivity of disclosing their smoking or alcohol consumption.  As an alternative, there are digital programmes either online or through mobile phones that support people to quit smoking and reduce their alcohol consumption.  But not many people know about these programmes and even fewer take advantage of them.

Digital programmes are successful in helping people quit smoking or reduce risky alcohol consumption.  As these programmes are available via websites or smartphone applications, they are available at little to no cost and no travel time.  They provide a private form of support for people who would prefer not to speak with their doctor or counsellor about these behaviours.  Under their One You campaign, Public Health England offers several health apps, which include a Smokefree app for quitting smoking and a Drink Free Days app for encouraging days off drinking.  

Our approach is to use the electronic health record to reach all patients who may benefit from digital programmes without adding any time or workload to healthcare professionals and with limited monetary cost to the healthcare system.  We will use information in the electronic health record to reach and refer all patients who smoke and/or drink alcohol at risky levels and use email or SMS to refer them to Public Health England’s mobile apps for smoking cessation and alcohol consumption.

Key research questions 

  • Objective 1: To establish the feasibility of using the electronic health record to identify adult patients who smoke and/or drink alcohol at risky levels.  We will also determine the characteristics of these patients and the availability of contact details for automated messages.
  • Objective 2: To explore the acceptability of receiving automated messages that promote smoking cessation and alcohol reduction apps, sent via the electronic health record, before and after messages have been delivered, framed by the Perceived Attributes of eHealth Innovations model, from both patient and staff perspectives.
  • Objective 3: To determine the reach of smartphone apps when promoted by automated messages sent via the electronic health care record.  We will also compare the characteristics of people who access the apps with those that do not.

Research Design and outputs 

This is a multi-phase, mixed method implementation research study:

  • Phase 1) Feasibility of sending automated messages via the EHR will be established by mining the EHR for the number and proportion of eligible participants, their characteristics and contact details;
  • Phase 2) Acceptability of sending automated messages via the electronic health record will be explored with qualitative focus groups before the messages are sent (pre-implementation), and with a quantitative questionnaire after the messages have been sent (post-implementation), both framed by the Perceived Attributes of eHealth Innovations model;
  • Phase 3) Reach of apps following promotion in the automated messages will be determined in a cross-sectional study design.

Settings: The study will take place at West Suffolk hospital, based in Bury St. Edmunds, Suffolk, UK.  West Suffolk hospital is an acute hospital that provides short-term care to patients.  It is one of 16 acute NHS Trusts that are internationally recognized providers of exceptional and efficient NHS care via world-class digital technology and information. 

Phase 1 (Feasibility): Electronic health records will identify:

  • a) Total number and proportion of adult patients who smoke and/or drink alcohol at risky levels;
  • b) Patient characteristics, i.e. gender, age, ethnicity, level of deprivation (Index of multiple deprivation) and chronic health conditions; 
  • c) Total number and proportion patients with contact information, i.e. email, phone numbers (landline or mobile), and patient portal access.

Phase 2 (Acceptability): We will ask patients their opinions on whether it is acceptable to use their EHRs to identify those who smoke and/or drink alcohol at risky levels and send them emails or SMS to refer them to Public Health England’s apps. 

Focus groups (before message delivery)

We will conduct three focus groups with patients who drink alcohol/smoke tobacco. We will also conduct three focus groups with staff: one with healthcare professionals who screen for health behaviours, a second with operational staff (i.e. admin, EHR team, communications team), and a third with senior IT managers.

Online questionnaire (after message delivery)

We will invite patients (who have received the message in phase 3 below) to complete an anonymous online questionnaire exploring the acceptability of receiving the messages.  Questions will also be based on the Perceived Attributes of eHealth Innovations.

Phase 3 (Reach): Absolute number (N), proportion (%) and representativeness of patients who access the app

We will use the findings from the Phase 1 to determine the format of the message, and the findings of phase 2 to determine the content and delivery of the message.

We will work with hospital-based information analysts and IT team to set up the automatic messaging system that identifies patients who currently smoke and/or drink alcohol at risky levels.  This system will be used to send an email or SMS that includes a link to either the SmokeFree or Drink Free Days app.

Outputs: We will publish our findings in peer reviewed journals and present them at conferences. We will also feedback the study results directly to patients and staff at West Suffolk Hospital. If the messages are successful, we can use the same idea to identify and refer patients who have other health problems such as poor diet and lack of exercise to digital programmes that support healthy lifestyles. The same system can be used in other hospitals across the United Kingdom and in other countries.

The research team 

  • Dr Zarnie Khadjesari (PI) and Tracey Brown (Senior Research Associate)
  • External:
  • Dr Lorien Abroms (Department of Prevention and Community Health, George Washington University, Washington DC, USA)
  • Dr Michael Amato (Truth Initiative, Washington DC, USA)
  • Dr Sherine Eltoukhy (The National Institute on Minority Health and Health Disparities, The National Institutes of Health, Bethesda, Maryland, USA)
  • Dr Henry Goodfellow (Department of Primary Care and Population Health, University College London, London, UK)
  • Dr Helena Jopling (West Suffolk Foundation Trust, Bury St. Edmunds, Suffolk, UK)
  • Dr Alex Ramsey (Institute for Public Health, Washington University in St. Louis, St. Louis, Missouri, USA)

Funding

Cancer Research UK