The graph is fitted with a polynomial that captures the apparent rise and fall of data. All other fits failed.
The crude data is as follows. Private hospitals are indicated, the rest are NHS.
| Hospital | Num Deaths | Num Operations | % Mortality |
| London Bridge Hospital (Private) | 0 | 1 | 0.00% |
| Harley Street Clinic (Private) | 0 | 3 | 0.00% |
| Manchester Royal Infirmary | 1 | 37 | 2.70% |
| Barts and the London | 7 | 129 | 5.40% |
| Hammersmith and St Marys Hospitals | 8 | 109 | 7.30% |
| Papworth Hospital | 27 | 250 | 10.80% |
| University Hospital of South Manchester | 5 | 45 | 11.10% |
| University Hospital of Wales | 13 | 116 | 11.20% |
| Southampton General Hospital | 29 | 241 | 12.00% |
| University Hospital Coventry | 7 | 58 | 12.10% |
| Wellington Hospital North (Private) | 1 | 8 | 12.50% |
| Freeman Hospital | 9 | 62 | 14.50% |
| University College Hospital | 12 | 78 | 15.40% |
| Castle Hill Hospital | 9 | 57 | 15.80% |
| Glenfield Hospital | 25 | 151 | 16.60% |
| Basildon Hospital | 16 | 95 | 16.80% |
| Royal Brompton and Harefield Hospitals | 25 | 144 | 17.40% |
| St George's Hospital | 17 | 94 | 18.10% |
| Victoria Hospital Blackpool | 16 | 83 | 19.30% |
| University Hospital of North Staffordshire | 15 | 75 | 20.00% |
| Bristol Royal Infirmary | 30 | 149 | 20.10% |
| Leeds General Infirmary | 18 | 86 | 20.90% |
| Nottingham City Hospital | 11 | 51 | 21.60% |
| Northern General Hospital | 13 | 59 | 22.00% |
| Morriston Hospital | 20 | 91 | 22.00% |
| King's College Hospital | 29 | 127 | 22.80% |
| Queen Elizabeth Hospital | 16 | 68 | 23.50% |
| Royal Sussex County Hospital | 13 | 55 | 23.60% |
| James Cook University Hospital | 21 | 86 | 24.40% |
| John Radcliffe Hospital | 27 | 109 | 24.80% |
| Liverpool Heart and Chest Hospital | 31 | 120 | 25.80% |
| New Cross Hospital | 10 | 38 | 26.30% |
| St Thomas Hospital | 31 | 105 | 29.50% |
| Derriford Hospital | 31 | 101 | 30.70% |
Obviously the Private sector hasn't submitted enough data to enable a comparison between public and private (which rather contradicts their idea of a free market and choice). But grouping the data together a clear trend is visible. As hospitals process more patients the mortality increases, until around 100 operations (3yr period) after which the mortality declines.
This is much as would be expected. Small hospitals can provide personalised care, particularly if those patients are individually paying. As number increase then individual care will suffer. However as hospitals become huge there are advantages from experience, well practiced procedures and more expensive equipment (economy of scale).
Because small hospitals provide less data there is also a larger error bar at the left of the graph. Average mortality is 17.6%. Harley Street which does only 1 operation a year (on average) would only expect a mortality every 6 years (every other 3 year study) if it was an average hospital and would need to maintain zero deaths for 16 years before statistics could conclude it was better than average (at 5% confidence limit)! Small private hospitals thus need to be studied over much greater time periods to assess their performance.
Hospitals to avoid however are the high scoring ones at the end which are also large hospitals and the data is much more likely to represent their actual performance. Performance should improve as patient numbers drop and staff can provide more individual care.
The final conclusion to draw from this is how it doesn't support an argument for privatisation. It suggests rather that while private hospitals might often provide better healthcare outcomes (but not always), they do so by only treated a very few individuals. A private system would then result in higher mortality in the population from people unable to afford the treatment. A public system should provide lower national mortality over all because all people have access to healthcare, even while having higher mortality on its books. Moreover many people arriving in public hospitals from very poor backgrounds may have complicating factors arising from their poverty that will end up on public mortality books, but private clinics will be able to avoid such factors by excluding the poor.
Public planners ought take heed also from the economy of scale that very large cardiac hospitals provide better outcomes than mid-range hospitals. Exactly as predicted by modern economic theory and the ideas of centralisation. While ideologically many people may not like this conclusion, ideology should always be fact based rather then prejudicial.
The obvious next approach is to combine this with budgets. I cannot see how private hospitals can provide more efficient healthcare given the low throughput. So already i suspect this won't support privatisation arguments either.
This is much as would be expected. Small hospitals can provide personalised care, particularly if those patients are individually paying. As number increase then individual care will suffer. However as hospitals become huge there are advantages from experience, well practiced procedures and more expensive equipment (economy of scale).
Because small hospitals provide less data there is also a larger error bar at the left of the graph. Average mortality is 17.6%. Harley Street which does only 1 operation a year (on average) would only expect a mortality every 6 years (every other 3 year study) if it was an average hospital and would need to maintain zero deaths for 16 years before statistics could conclude it was better than average (at 5% confidence limit)! Small private hospitals thus need to be studied over much greater time periods to assess their performance.
Hospitals to avoid however are the high scoring ones at the end which are also large hospitals and the data is much more likely to represent their actual performance. Performance should improve as patient numbers drop and staff can provide more individual care.
The final conclusion to draw from this is how it doesn't support an argument for privatisation. It suggests rather that while private hospitals might often provide better healthcare outcomes (but not always), they do so by only treated a very few individuals. A private system would then result in higher mortality in the population from people unable to afford the treatment. A public system should provide lower national mortality over all because all people have access to healthcare, even while having higher mortality on its books. Moreover many people arriving in public hospitals from very poor backgrounds may have complicating factors arising from their poverty that will end up on public mortality books, but private clinics will be able to avoid such factors by excluding the poor.
Public planners ought take heed also from the economy of scale that very large cardiac hospitals provide better outcomes than mid-range hospitals. Exactly as predicted by modern economic theory and the ideas of centralisation. While ideologically many people may not like this conclusion, ideology should always be fact based rather then prejudicial.
The obvious next approach is to combine this with budgets. I cannot see how private hospitals can provide more efficient healthcare given the low throughput. So already i suspect this won't support privatisation arguments either.
