IMPROVING DATA QUALITY
The client is a large English Academic Medical Centre. They provide secondary care to one of the oldest populations in England and are also the provider of specialist stroke, heart-attack and cancer treatment to a population of circa 1-million; their cancer services reflect the demographic served in that they rank as the fifth-largest cancer center in England.
The client operates from a base of approx. 1100 beds and has a revenue budget of £640m per annum.
At the time CER were engaged, the hospital was under intense scrutiny and high degree of regulatory intervention. Despite managing to balance quality, patient access and finances successfully for a number of years, a rising Hospital Standardized Mortality Ratio (HSMR) coupled with a significant predicted financial deficit; at that time in excess of £25m for the year 2015/16. Achievement of access targets had also become erratic.
CER Context & Support
The hospital were members of the Association of UK University Hospitals (AUKUH). CER had been the designated provider of benchmarking services to the AUKUH for years and the client hospital had previously been an active participant in the AUKUH benchmarking club.
A refresh of members of the hospital senior leadership team led to the newly appointed Chief Operating Officer (COO) reaching out to CER. Specifically, to see if, through our curated and facilitated benchmarking service, we could help highlight opportunities to both better explain and so improve the hospitals performance. Of greatest concern in the brief was the seemingly ‘hard to explain’ rise in HSMR (1.15:1 and rising).
The CER team undertook an exercise with the COO to:
Explore in detail the demographic context of the hospital, including the relative deprivation overlaid at a granular level with activity profiles
Contrast detailed population and service profiles with carefully curated peers
Identify immediate areas for improvement
From this initial exercise it was clear that the hospital either had a unique population of elderly patients i.e. the data suggested few if any had the sort of complex co-morbidities expected in patients of this age group or there was a significant issue with data-quality (capture and episode / diagnosis coding)
By providing the hospital with a detailed breakdown of the services and clinical areas most out of synch with their peers they were able to implement a significant data capture and coding improvement scheme. The main outcomes were:
1. Improved patient and system confidence in the safety of the services provided
Through improving the capture of patient episode data, a much more accurate and detailed diagnosis coding was possible. Fig. 1 highlights the rise from a mean diagnosis (all-patients, all episodes) of approx. 4.5 in 2015 to 6.0+ in 2017/8
Fig 1. Mean Diagnoses-per-Episode
As more accurate and appropriate profile of the patients being treated by the hospital was captured. The relative risk of what was clearly a higher-risk, multiple co-morbidity patient population was more accurately captured and as such a much more accurate HSMR was reported; at the time of writing the HSMR had fallen to lower than expected at 0.97:1 and falling (Fig. 2)
Fig 2. HSMR
2. Improved Income Performance
Whilst not the main driver for the client, improved data capture and coding produced a significant improvement in income; the correct capture of co-morbidities and diagnoses increasing income by circa £8m in 2016/17 and an additional £7m (£15m total) in 2017/18