Our new report measured the productivity across healthcare organisations and found that significant variation remains, write Andrew Street and Adriana Castelli

While the NHS as a whole has delivered recent productivity growth over successive years, beneath the national level there remain significant variations in productivity across healthcare organisations.

Eliminating poor productivity will be key to securing future productivity improvements.

The latest publication from the York University’s Centre for Health Economics focuses on productivity variations across NHS hospitals, applying techniques consistent with how productivity is measured at a national level.


Hospital productivity is measured as the ratio of outputs to inputs. Outputs capture the quantity and quality of care for hospital patients. Inputs include staff, equipment and capital resources used in providing patient care.

Routine activity and financial data are used to measure the output and input of all English hospitals in 2008-09 and 2009-10.

We found find substantial variation in hospital productivity: the 10th and 90th percentiles being 16 per cent above and 16 per cent below the national average in 2008-09 respectively, and 14 per cent above and 16 per cent below in 2009-10.

‘Trusts created as a result of mergers realised improved productivity’

Productivity scores were not particularly sensitive to whether or not quality of care was included in the assessment.

The position of individual hospitals did not vary greatly from one year to the next, though notably both trusts created during the period as result of mergers realised improved productivity, these being South London Healthcare Trust and Western Sussex Hospitals Trust.

The bottom line

We found that productivity was lower in hospitals with greater financial autonomy.

The difference was driven by the higher expenditure on capital by foundation trusts. This implies that investment by foundation trusts in their infrastructure takes a while before it yields proportionate increases in output.

Productivity was also lower in hospitals where a large proportion of income derives from education, research and development, and training activities. This may reflect cross-subsidisation between patient care and teaching and research activities.

The data predates the Department of Health’s major review of the service increment for teaching.

We also found that hospitals treating greater proportions of children or elderly patients also tend to be less productive. This may reflect the set of 1,400 healthcare resource groups used at the time being unable to adequately capture case mix differences among hospitals.

The new set of health resource groups, called version 4+, which is now in use may have resolved this.

Lessons learned

What are the lessons can we learn from this variation in productivity?

One is that variation persists in the NHS. In 1989, the Working for Patients report highlighted variations in practice as a rationale for fundamental reform of the NHS; in 1997, The new NHS: modern, dependable white paper adopted a similar theme; in 2002, it was announced that payment by results would tackle cost variation among hospitals.

But the financial regime has not eliminated variations – as evidenced in this study – and needs to be strengthened in two main ways to make inroads into the problem.

First, payment by results does not seem to offer sufficiently strong incentives to encourage hospitals to improve their productivity.

The tariff is not as effective as it might be, with hospitals able to offset tariff losses with income from other sources – a strategy dubbed “tariff leakage” in Monitor’s annual planning review guidance.

Second, incentives should be sharpened to place greater pressure on the least productive to improve. An across the board 4 per cent efficiency requirement – as in the 2014-15 national tariff payment system – actually does the opposite.

In more productive hospitals further improvements are harder to achieve so an across the board target unduly penalises these organisations.

In responding to such incentives, hospitals need to scrutinise their relative productivity and cost structures when making decisions about their scale of operation. It is not clear that they do so in a systematic fashion.

The last few years have witnessed substantial increases in the amount of activity undertaken in hospitals. This may have been fuelled partly by hospitals believing that expansion is profitable, with national tariff exceeding any increase in additional costs.

‘Hospitals need to consider the implications of changing the level and mix of activity on both revenue and costs’

But this belief may be mistaken. Expanding activity might necessitate taking on more staff or building more facilities. If these additional costs exceed tariffs, hospitals run the risk of accumulating a deficit.

Hospitals need to consider the implications of changing the level and mix of activity on both revenue and costs.

The reference costs don’t allow them to easily do this; in future, patient level costing systems might offer greater insight.

It is also instructive to consider the nature of increased activity. There has been much debate among commissioners about preventing or reducing unplanned admissions.

But, in fact, it is elective activity that has accounted for most of the growth in hospital activity. In the 14 years between 1998-99 and 2011-12, non-elective activity, including maternity, increased by 21 per cent.

In contrast, over the same period elective activity increased by 66 per cent, and not all of this will have been due to pressure from reducing waiting lists or from population growth. Clinical commissioning groups will have to manage both elective and non-elective activity in order to live within their means.

Andrew Street and Adriana Castelli work for the Centre for Health Economics at York University. The report is an independent report commissioned and funded by the Policy Research Programme in the Department of Health. The views expressed are not necessarily those of the DH.