In the early 1990s, health and social care organisations knew surprisingly little about the actual costs of services, particularly unit costs such as the cost per working hour for a social worker, the cost per hour of face-to-face time with a service user, or the cost per GP consultation. More often than not, rough estimates would be arrived at by simply dividing known expenditure by activity, without any analysis of the different components that contributed to the overall cost. When researchers needed appropriate health and social care unit costs these would be calculated on an individual study by study basis, with no generalised source of information for everyone to draw upon.   

 

This absence of information belied the growing demand for cost data by policy makers, commissioners of services and providers. My colleague Jennifer Beecham and I had done some work on the theory and practicalities of costing community care. But when the Department of Health said it would like an up-to-date understanding of unit costs, my first reaction was that it was not possible.

 

However, after thinking through the sort of information that was available from previous work and current sources, we came up with the idea of creating a ‘schema’ that pulled together the different cost elements – the ‘building blocks’ of cost information –  that could be identified for a specific type of service. This approach was both transparent and flexible. It meant that users of health and social care unit cost estimates could adjust the information that was fed in to produce the estimates, thereby ensuring that the numbers would reflect the particular context and purpose for which they were intended.

 

I was fortunate to lead the work that produced the first Unit Cost of Health and Social Care volumes in 1993, assisted by Steve Smart who was seconded to PSSRU from the Department of Health. The initial step was to identify the key elements for constructing a unit cost, such as client group, type of service, provider agency, resource type and type of service. The main challenge at the start was to obtain the basic information as there were so many gaps in what was known. We drew on other PSSRU work, including various individual studies and comprehensive cost analyses led by Martin Knapp and Jennifer Beecham. We collated information from a wide range of sources, bringing together, for example, data on actual salaries, how they changed over time and existing research on how health and social care workers spent their work time. We aimed to provide readers with a set of cost schema containing specific information about the cost of each service covered; a commentary detailing the basis for the estimates; price indices; a reference list of key studies; a glossary and indexes. There were of course many challenges – particularly the lack of data, changing services and problems in defining units – but with the first volumes we demonstrated that the general approach worked.

 

Further impetus for the work came from a Department of Health request for a ‘ready reckoner’ for health service costs that included the investment that went into producing a qualified health professional. Thus the unit cost of a doctor’s time could not simply be seen as their salary, but needed to include the cost of getting them qualified. This approach was later extended from health to social care workers.

 

We could never have known at the beginning but the Unit Cost of Health and Social Care volumes have been published every year since 1993. It was quite a surprise to me the first time I went to a health economics conference and our work was so warmly received. The volumes saved a lot of people a lot of time, and quickly became one of those things that researchers relied upon. Over the years we worked each year to expand the range of services covered and to improve the basis for the estimates. Our advisory Working Group of civil servants and academics, which meets once a year, has been invaluable in identifying priorities and informing us about new and current work and sources that the analysis can draw on.

 

I led the work while it formed an element of core Department of Health funding for PSSRU. After PSSRU core funding ended in December 2011, Jennifer Beecham took over the lead on a separately funded stream of work. She was also responsible for obtaining Department for Education funding for children’s social care, which was dropped for a while when responsibility for these services moved from the Department of Health. Whenever there has been any question about ongoing funding there has been quite a lot of pressure to keep the unit costs publications going.  

 

These days the compendium is downloaded in part or in full at least 600 times each month. The data are used by central government departments, local authorities and have had a major impact on research: more than two-thirds of health and social care economic evaluations in England published between 2008 and 2012 used the unit costs to underpin their analyses. The volumes now cover unit costs for more than 100 health and social care services. The area in which it remains hardest to obtain good information relates to people’s use of time – how much time is spent by health and care professionals with individual patients/service users, and how much additional time must be spent by the same professionals to facilitate that time with patients/service user. 

 

Jennifer Beecham now oversees the unit cost publications with Lesley Curtis, who has for a long time acted as the main researcher. The work continues to evolve and one recent improvement has been to keep a closer eye on the relevant literature and to be more proactive in approaching researchers for sources of information.  In other countries there is quite a lot of interest in us potentially creating similar data for them, but I think quite a lot of the time they just use ours. I don’t think any other country has yet done what we achieved on unit costs.

 

Interviewed by Teresa Poole