Five simple rules for running a first-class hospital
Oct 21st 2010
MEASURING good health, in patients and hospitals alike, is one thing. Finding the causes is harder. Medical professionals in Britain know that Hammersmith Hospital in west London, for example, is one of the best in the country—Tony Blair received treatment there for his irregular heartbeat and it is one of only a handful of places to which London ambulance crews take people with suspected heart attacks. From America to Sweden, the best hospitals in a rich country outperform the rest. But how?
Stephen Dorgan of McKinsey, a consultancy, and John Van Reenen of the London School of Economics have tried to answer this. They studied almost 1,200 hospitals in America, Britain, Canada, France, Germany, Italy and Sweden, using techniques more commonly applied to identify excellence in manufacturing industry.
The hospitals with the best management practices (analysed as if they made things rather than curing people) also ranked best on a standardised measure of medical success: death rates among emergency patients experiencing heart attacks. That score works across countries and cultures, and has unambiguous results.
The researchers found five characteristics associated with the management of successful hospitals. One was competition—or at least the perception of having competitors. Hospital managers who named more than ten institutions that they competed with scored more highly on their management practice than those who saw fewer alternatives for their patients to choose from.
Having lots of small providers vying for patients will not on its own raise standards. The researchers also found that bigger is better when it came to good management. Hospitals employing 1,500 or more staff are better run than those employing more than 500, which, in turn, outperform those with more than 100 staff. Hospitals with less than 100 people working in them are particularly badly managed. Private ownership is another factor helping hospitals score more highly.
Only in Italy must hospital managers have clinical degrees. That seems a good rule: institutions that employ clinically qualified staff in management score better than those that do not. Overall, Britain performs badly by this measure: in Sweden, 93% of hospital managers are former doctors, nurses or other clinical staff. In America, Canada and Germany, the proportion is 71-74%. In France it is 64%, and in Britain just 58%. Good staff also need the freedom to exercise their own judgment: managers with the most autonomy fare best.
That may be a signpost for improvement. Hammersmith Hospital is one of five such institutions run by Imperial College, London. Stephen Smith, who is responsible for the hospitals, is also principal of the faculty of medicine. In America, Johns Hopkins is not merely a highly rated university based in Baltimore, but its medical arm also runs a hospital of high repute. Big, professional, autonomous, mostly independent and ferociously competitive: elite universities seem to offer all five important characteristics for saving lives.