Drowning in bad data

I really dislike the phrase: “What’s measured is managed.” What I hear is: “What isn’t measured isn’t managed.” And what’s measured badly is managed worse.

Unfortunately we are drowning in bad data.

Let’s say the CEO of the UL Hospitals Group is in front of the Joint Oireachtas Committee on Health. They ask why Ennis has twice as many inpatient discharges as Nenagh, and a much shorter length of stay, although both are 50-bed medical units.

The answer is not in the data.

I’ve seen the local figures. I know that “inpatient discharges” now (bizarrely) includes people sent home from assessment units. Ennis has an established and busy unit, while the staffing for Nenagh is being set up.

Yes, they are a valuable service to patients and GPs, but we used to call these cases “urgent outpatients”. It makes no sense to add them to the patients who need ward level care, to create a composite length of stay.

If the Committee ask the CEO about waiting lists, there’s plenty of data. But if they ask about patients on trolleys in University Hospital Limerick (UHL), there’s nothing.

I know. I spent my holidays last year researching the situation. It is not an IT problem. It is the rules of HiPE.

There is no data on age or illness, whether the trolley patients were referred by GP or transported by ambulance. The CEO can’t even tell the Committee which consultant team has admitted the patients.

Until the trolley is physically pushed out of the emergency department, there is no record of the admission. A patient treated for two nights on a trolley, then two nights in a ward bed, is recorded as a two-night admission.

Most trolleys are medical cases, so length of stay figures for medicine are completely weird and should be ignored.

The only reliable item on trolleys is the number, but the HSE fudges it.

Every morning at 8am, the Irish Nurses and Midwives Organisation (INMO) counts the number of patients admitted on trolleys. Then the HSE tells us blandly that only half of them are there more than nine hours.

Rubbish! If you are admitted and on a trolley at 8am, you have already been there all night.

Because there is no proper data the national conversation on trolleys remains confused. Because trolley admissions are not measured, they are not managed.

The CEO could point out that a key cause of the trolley problem in UHL was the reconfiguration of the mid-west hospitals, which they inherited.

Together, Ennis and Nenagh used to admit up to 200 ward patients. Now that is cut in half. Those beds were not replaced in UHL.

Actually, there’s lots of very bad data on acute hospital beds.

In 2002, the Department of Health issued an excellent report (introduced by then Health Minister Micheál Martin), which recommended that we needed 15,000 hospital beds.

But in 2007, the HSE commissioned a management consultant report, which said only 9,000 beds were needed. I didn’t read the report at the time and I’m pretty sure no one else did.

When I eventually read all four volumes, I stared in astonishment at Appendix K, which outlined that the bed number includes 3,000 day beds. If we had understood back then that this report recommended only 6,000 inpatient beds, there would have been uproar and possibly revolution.

In addition, from talking to people involved in the report, I realised that hard conclusions were drawn from fluffy data. We would have laughed at the report if we’d known.

Instead we all sank into depression and a sea of trolleys.

The HSE has been closing beds for years, apparently based on this report.

At least the CEO can rely on some of the other numbers. The Health Committee may be surprised to learn that UHL treats more inpatients than the Mater, has more ED attendances than St Vincent’s, and that there are over 2,000 employees on the main hospital site alone.

Some commentators think Dublin hospitals are big and everywhere else is rural and small.

Then there are the voices crying out that Ireland has the same population as Manchester and should run its health service the same way. Baloney!

A recent EU report says Ireland has one of the most rural populations, 40 per cent as compared to 12 per cent in Britain. And the Lightfoot report on the ambulance service explains that this “high rurality” is a major problem. Although Irish people call ambulances far less than in England, the service struggles because travel times to the patient and to hospitals are so much longer.

No wonder we worry when emergency services are moved further down the road.

It is time to drop the bad data and pay attention to the good stuff.