Andrew Heed, B. Pharm. CPIO, Integration Architect. The Newcastle upon Tyne Hospitals NHS Foundation Trust. Afroze Mobasher, Pharm. D Analyst Configurer Imperial College Healthcare NHS Trust
Local reflection of an optimisation project Optimisation cycle Collaborative optimisation Future optimisation challenges.
ImplementedCerner Millennium in 2009. Electronic prescribing and administration PAS Theatre module A&E Electronic orders. Not live with: Problems and diagnoses Documentation Certain medication e.g. Insulin. Small scale optimisation steps.
Anti-microbial stewardship: Simple messages to reduce cephalosporin use linked to improvements in prescribing in ED and over 65 yo improvements in MRSA rates. Specific order sets Acetylcysteine for Paracetamol overdose Hyperkalaemia rules and order sets. Work best when designed with users. AKI algorithm and alerts. Interaction rules. Parenteral nutrition ordering.
Optimisation cycle: Measure Analyse Investigate Design Control Test Engage Implement Reflect Exploit Collaborate You fill in the gaps You fill in the gaps You fill in the gaps You fill in the gaps You fill in the gaps
Hospital Scores for detection of test orders causing an ADR according to product Metzger et al. Mixed Results In The Safety Performance of Computerized Physician Order Entry. Health Affairs 2010 29(4): 655-663
epx system Central drug catalogue Standardised nomenclature Central and configurable routes of administration, frequencies, units of measure Order sets. Order sentences. Legibility. Reporting. Tool for change.
Initial build not good enough : No electronic blood glucose results. Still needed two charts. Unable to view drug doses in lab view. Ability to search by brand/ type. Couldn t lock down unit of measure or route. Staff were new to the system. Skill gaps on my part. Not going live was the correct decision.
Experience gained from smaller scale optimisation. What works well and when Build re-thinks Rules, reporting, other system build Rapport. Multiple approaches Experience from others. IA course, discussion groups, UCERN, sign up to safety. Electronic blood glucose results.
MDT design group: Baseline data from annual audit / datix reviews Specific Insulin build and views. Automated pharmacy referral. Automated referral to DSN. Specific targeted build Humulin R. Reporting tools. Go-live Used reporting to target change and review.
Electronic Glucose Monitoring View A colour coded view of laboratory results, insulin doses, oral hypoglycaemics, glucocorticoids. to highlight out of range values.
Table 1. Inpatient Insulin prescription error types occurring pre and post the introduction of electronic insulin prescribing in Newcastle
Different types of error.. Dose not signed. Open-ended prescriptions. Doctors no longer required to review. The order requires a dose. Review task for doctors. Based on experience with pharmacy.
So if we can use the epx system to do all that could we...
Based on data from Epx: Unable to see all patients. Do we need to see all patients? Do we see high risk patients. Can we optimise the system to prioritise clinical team? Direct reviews Quantify service need and service delivery. Move away from location based service to a problem / patient-based service. Better value with targeted service?
Implemented a pharmacy task list Based on requests from clinical team System rules generate a follow-up task attached to patient record. Some data collection forms. Initial good response Limited by volume of work and cluttered views Teams continued to be ward based Feel like we have to clear the list Good system but some design flaws.
How to control: Optimisation requests, template requests Ownership and review How to test and monitor Workload and maintenance Drive awareness of possibilities Design principles The simpler the better vs complex over specification Promotion and education
Safety case and closure report: Is it needed? Who is responsible? Standard change vs complex redesign How does this affect other processes Does it change the system from vendor perspective. Training
Shelford Group collaboration Sites using Cerner EPR Varying experiences and scope. Experiences of optimisation by prioritisation. Brief: Design and develop a multi-system, real-time, dynamic process to create a medication Acuity Score to enable prioritisation of clinical pharmacy teams.
Cross-site consultation on acuity Scenarios Reviewed by EPR team. Classified by type. Classification Count Antimicrobial Stewardship 17 Therapeutic drug monitoring 12 drug-drug interactions 11 Insulin 11 Anticoagulants 10 Location Based Risk 9 High risk drug / class 19 Patient factors 20 Total 109
Scenarios reviewed and assigned a rule template Template type Number Order based. 36 Result based. 21 No template applicable. 18 Interaction 15 Clinical event with order 13 order detail - route of administration 6 order detail - location 5 Order event with co-factor 4 Admit rule 1
Considerations: Complexity of rule: Can it be done, time to build, Local variations in scope and content. Workflow. Several options for rule output. Task based Dynamic vs Static. Scoring system. What does the score mean.
30+ Rules in draft status. Approx.10 templates developed: Admission rules High risk order (unchanging) High risk order (changing) TDM rules (changing) Others e.g. weight.
Scenarios: high risk drug A or B, complicated by age over 70 and or low body weight. Specific rule model: Requires the following: Separate rule model Requires the following: Drug A rule Drug A + patient over age 70 rule Drug A + patient over age 70 + low body wt rule Drug A + low body wt rule Drug B rule Drug B + patient over age 70 rule Drug B + patient over age 70 + low body wt rule Drug B +low body weight rule Drug A rule Drug B rule age over 70 rule, low body weight rule
Developed rule base at Newcastle. Exported to colleagues at Imperial for localisation and testing. Review build and agree final model. Collaborate with The Newcastle University School of Pharmacy. Evaluate the Acuity model. Silent Live period. Review vs current practice. Define visualisation.
Clinical Informatics literacy within healthcare. Many sites not using Many sites have low engagement Major barrier to communication on both sides Both guilty of jargon Is what we get truly optimal Do we understand data and structure Can we get informatics on undergraduate and post-graduate teaching?
Better prepared for using systems in practice Opportunity to think differently Secondary uses of the system Management and use of data System development Awareness of limitations of the system
Afroze Mobasher
Lithium Anticoagulants Clozapine Methotrexate Parkinsons medication Variable dose insulin Insulin infusions Digoxin NOAC Immunosuppresants Inotrope Chemotherapy
Hook was broken, importing the rules adds @quot; before and after the LOGIC piece which needs to be corrected for each rule
Competing pressures Shared domain: Chelwest meds go live 2019 Resources for process harmonisation Lloyds pharmacy paperless workflow PGD workflow
Prior experience of: EKM rules and Discern Expert Meds Order Catalogue architecture DTA build and design iview build and design Frontend PowerChart workflows Recommended Cerner courses: B&M Core Foundations B&M Documentation Management B&M iview B&M Medication Processes EKM rules
Aligning imported meds order catalogue to local Synonyms Encounter type Encounter status Frequencies DTA build Hook logic iview Build Identify locations: ICU, chemotherapy etc Testing and Validation: EPMA Pharmacist
Standardised nomenclatures: Enables core rules with less localisation. Multum Flip - dm+d terminology to enable use of VTM terms where applicable. Snomed CT for problem based rules. Standardised laboratory nomenclature. Vendor assistance May need aliasing option to drive nomenclature? Creation of a package to install rules A wizard to automatically map
Thank you for your time