Why were ESTEEM findings so different?

The results of the ESTEEM trial were so different from our practices’ exparience that we wanted to work out why.  We compare the design, measurement and outcomes here.

InterventionESTEEMGP Access
CONCLUSIONIncreases workloadAs a system, it works
DesignCluster RCTDiverse, self selected over 3 years
EthosStandardiseNeed to adapt continuously
Subject motivationTake part in research, paid for triage session time.Desire to improve service, save time.
Term4 weeks run in, 5 weeks capture12 week change programme. Sustained indefinitely. Variable data series
Size13 in GP triage arm, 6,658 consultationsData from 66 practices, 1,600,000 GP phone consultations
Demand"Same day" requestsWhole system, all demand
TrainingGP only?Whole team
DemandNoYes, month/week/day/hour
Response & waiting timesNoYes, minutes to phone call, days wait to see GP, before & after
ContinuityNoYes. Three methods, UPC from records, patient preference, GP preference
Consult durationYes, GP recorded manuallyYes, direct from clinical system, phone and face to face
Resolve rateCan be derivedYes
2ary care impactYesLimited
Patient satisfactionPostal survey at 4 weeksTelephone survey same day
Feedback to practicesNoneWeekly during intervention
Change leadership and staff survey?Yes, before and after intervention
Phone resolve rate25%60%, range 50-70%
GP productivityDownUp around 20%
Response time?Range 20-80 minutes
Wait to see GP?Down 80%
Continuity?Range 0-20% higher
Patient satisfactionLittle change70% say better
20% same
10% worse

For a fascinating read on another RCT study with a similar tale see “Did the trial kill the intervention?”  which ends with Findings from this study strongly suggest that within the local environment where the trial is conducted, acquiescence from those in positions of authority is insufficient; commitment to the trial is required.”

Leave a Reply

Your email address will not be published. Required fields are marked *