Conporto spreads its safety net to catch patient harms before they happen
It sounds like a health sector version of the film Minority Report, a computerised “precrime unit” capable of detecting possible harms to people before they occur.
The Conporto event detection and mitigation system is, in fact, an automated tool that analyses patient records to detect potential medication-related harms.
Incidents are flagged to the GP’s inbox in their PMS.
Patients First chief executive Helmut Modlik says the approach is “revolutionary”.
“It’s the first of its kind in New Zealand and, as far as we know, in the world,” Mr Modlik says. “One of the biggest problems in health worldwide is the increasingly digitised silos of health data. Interoperability across these datasets is the Holy Grail.”
Conporto Health Ltd is the trading arm of Patients First, a not-for-profit charity owned by General Practice New Zealand and the RNZCGP.
Mr Modlik says the motivation for the Conporto software came after research into medication-related harm, in this country and overseas. Studies suggested nearly a third of harm events originate in primary care before a hospital admission.
In partnership with health IT company DrInfo, Patients First decided to develop software that could be installed on servers hosting patient databases, and scan them for possible harm events.
In March, the system was tested in 94 general practices and 152 pharmacies around New Zealand. More than 52,000 medical appointments and 137,000 repeat prescription reminders were monitored, detecting 196 medical risks of harm and 136 pharmacy risks.
These included inappropriate prescription of NSAIDs, metformin or allopurinol, to patients with renal impairment, or prescription of an antibacterial macrolide with a statin. There were 136 prescriptions of sodium valproate for 16 to 45-year-old women without a contraceptive prescription.
Results showed 100 per cent of harm-event notifications were successfully sent, then opened and viewed by GPs.
Mr Modlik says the trial proved the system can be used by doctors without any specific training, and fits smoothly into their normal workflow. “It’s business as usual, but with high-value extra information at point of care.”
GPs easily adapt to system
GP Chris Reid’s Kerikeri practice was involved in the trial, and Dr Reid is on Conporto’s clinical advisory group. He says it was interesting to see how easily GPs adapted to the system, opening the notifications in their inboxes and taking action, often discussing issues in their morning “huddle”.
Some GPs might feel there is a slight Big Brother aspect to it, but he says it’s good for patient-centred care.
“I’m keen for us to use this, and be seen to use it. If we can prevent harm, then not to do it is almost unethical.”
Mr Modlik says there is no human involvement with the detection system; it views and queries data without extracting anything or creating a repository, thereby minimising privacy issues.
The team is running an extended trial until the end of October, refining the system.
There is “no logical limit” to the records that could be analysed, Mr Modlik says. Conporto is talking to the Ministry of Health, ACC, DHBs and other organisations such as the Health Quality & Safety Commission, Pharmac and the RNZCGP on how to scale-up the system before the end of the year.
Software trawls data in search of harm
The Conporto system finds a patient’s NHI number across the various networks, looking for data points that relate to specific harm risks.
Patients First and DrInfo have coded data points for nine conditions which they say, if not identified, mean patients are “virtually guaranteed to end up in hospital or worse”.
If enough data points relating to a harm event are identified, the clinician is sent a notification at the start of the day, with a brief narrative description signed off by an independent clinical governance group.
The clinician then receives a further notification in their PMS at the time of the patient’s appointment, giving them access to an integrated patient summary record with all the relevant database information in a single view.