Guardian report on HES data

The Guardian had an interesting story yesterday on the poor quality of HES data:
http://www.guardian.co.uk/society/2010/jun/13/nhs-statistics-flawed
Includes discussion of data in surgery, comparing that collected by surgeons with that collected via HES route by administrators. Also mentions how data has been used to move complex surgery to specialist centres, and the subsequent impact on patient safety, covered in more detail in an accompanying article:
http://www.guardian.co.uk/society/2010/jun/13/specialist-surgery-increases-survival
The research comparing HES and surgeon-collected data (for abdominal aortic aneurysms) is also shared:
http://www.guardian.co.uk/news/datablog/2010/jun/13/abdominal-aortic-aneurysm-surgery-statistics

Bits and bobs

Project Manager Today, June 2010

Interesting article on project to renovate Lincoln Centre for Performing Arts in New York – involving New York City Ballet and New York City Opera – had to fit around performances. Highlights the need for clear governance and decision-making (a number of committees focused on specific specialist areas which all had to escalate decisions for approval to an overarching committee so all decisions signed off in same place; need for constant, ongoing engagement with those affected to minimise disruption and to promote understanding of rationale behind decisions, and to enable shared decision making where possible)

Joseph Czarnecki includes an interesting comparison on PMBOK v PRINCE2

Henny Portman writes about avoiding PRINCE 2 In Name Only – must follow 7 principles of PRINCE2 to do properly – continued business justification; learn from experience; defined roles and responsibilities; manage by stages; manage by exception; focus on products; tailor to suit project environment.

Dante Peagler writes about the need for effective planned project handovers – so often, there is no formal handover and there is then a period where the new PM has to catch up and may not pick up on the tacit knowledge. Need to allow enough time, ensure project documentation is up to date, introduce new PM to key stakeholders, work in parallel if possible, make transition dates clear.

HSJ 27/5/10

Interesting article on risk modelling - which includes mention of PARR and West Mids BUPA solution – and growing use of virtual wards to avoid A&E admissions. Now looking into social care applications – identifying elderly people who are at risk of losing independence through ageing/ill health. 80% of PCTs estimated to be using predictive modelling. Different approaches to virtual wards – nurse-led with GPs becoming involved as required (Croydon), dedicated virtual ward doctors (Wandsworth); general practice led (Devon); virtual discharge ward running predictive model on inpatients and offering those at high risk of readmission 30 days of support (Toronto). NHS Devon now looking to commission a front end – including dashboards – to help manage their GP led service.

Also supplement on Health Intelligence focusing on PHOs. Outlines following challenges:

• smarter management of multiple data sources
• helping local users understand and use health intelligence in decision making
• intelligible health information for informed consumers
• more intelligent approach to real time data
• filling persistent gaps – new health and quality measures and reliable population-based registers
• better sharing of health and local authority data
• better grip on measuring the whole pathway of intervention.
  

Talks about the tension between needing data quickly or needing assurance that data is clean – (reminds me of usefulness = validity x relevance/work). Also talks about role of PHOs e.g. production of annual health profiles, and topic-themed PHOs e.g. obesity, injury, child and maternal health. NWPHO has alcohol profiles which are being used across the NHS. London PHO has developed health inequality tools – smoking cessation, statin prescribing, antihypertensive prescribing, controlling high blood sugar in people with diabetes, interventions to reduce infant mortality. Also considers how PHOs support WCC showing example of Y&H’s SPOT tool, to help prioritise investment and disinvestment. Also mentions the SHAPE (Strategic Health Asset Planning and Evaluation) toolkit to inform strategic planning of services and physical assets. Nice to see mention of NLPH.

HC 2010 notes

Hmmmm....some time since the HC2010 conference but finally here are my notes:

- Gwyn Thomas – Informing Healthcare talked about the programme in Wales. He covered how they established “rules of the road” at the start to manage communication and expectations. The approach has been to take mostly small rapid steps with occasional large ones. There has also been an emphasis on learning together and developing a sense of common purpose. The architecture model considers different types of architecture: information; systems; and social (trust, relationships, collaboration).

- Aidan Halligan – gave a talk on values. He quoted Don Berwick “culture eats strategy for breakfast” giving the example of what hasn't worked with the NPfIT. He outlined how the explosion of knowledge has outstripped the ability of the NHS to keep up and some of the tools they are using in Imperial such as After Action Reviews (lessons learned debriefings) but set up to ensure that lessons learned are immediately implemented.

- Richard Hamblin from CQC – talked about the new Quality and Risk Profile and how it will show the synthesis of information on an organisation and will be used to identify problems. V1.1. available May, v2 in Autumn. A network of RIEOs has been set up (Regional Intel Officers). The plan is not to publish risk estimates but they do plan to publish judgements of inspectors. Data is fed and aggregated into a risk model – this is used to produce a dashboard which has dials (RAG) prepared by analysts – the dashboard is shared with the organisation; the inspector will then confirm any red lights (or not) at the inspection. Dashboard measures are around involvement, innovation; quality and management; suitability of staffing; safeguarding and safety; personalised care, treatment and support; suitability of managementt. Drilling down enables exploration of inherent risk, situational risk, population risk, uncertainty risk. Indicators are derived from HES, patient/staff surveys, national clinical audits, regulators and bodies such as PEAT and NPSA. Publicly accessible data will be based on the inspection not data alone – public will be able to view high level dials and be able to drill down into inspection report. Planning to provide access to principal commissioner.

- Samantha Riley – SEPHO – presented on how they have changed the culture of presentation and visualisation of data and information, through education/training and design of dashboards. They now provide regional benchmarking for 6 standard indicators (example of incorporating knowledge from library service on MRSA/C Diff). They're looking at how to evidence variation e.g. one example around hip replacement was due to innovative practice – needs to be shared to improve length of stay, mortality, readmissions across the health economy.

- Robert Lake – talked about Strategic Programme for Adult Social Care Information – includes development of the Adult Care Support Record ( aCSR), review and update of social care information guidance, zero-based review of data/information needs. 5 key themes – standards, aCSR, intelligence (NASCIS launched by IC in Oct 09) and JSNA, infrastructure, workforce (culture, leadership, training). NHS number to be applied to all adult social care records. aCSR will lead to a NMDS and data dictionary. Need to engage with self funders – currently no information on this, not even sure of extent of population. aCSR will assist personalisation agenda, by helping individuals to identify and manage their needs.

- James Walker – talked about localising MoM. Suggests that a typical pathway involves 800 hours of development. Highlighted importance of clinical ownership. Demonstrated ability to drill down to BNF and NHS Evidence. Outlined differencess between administrative and clinical localisation and implications of each.

- Rick Jones – talked about national pathology info strategy. Data in pathology has unpredictable downstream uses, can be merged from multiple sources and persists over time. Tests throughout 18 week pathways – data collected from different places (e.g patient in Cumbria could get referred to number of DGHs) and there are some variations which may not be taken account of if the data is used later for R&D. For example, different formulae can be used to test for chronic kidney disease – can lead to misclassification, unreliable results, underdetection, inappropriate referrals. Equivalent of BNF for lab tests is run by Royal College.

Business intelligence system at Shrewsbury

E-Health Insider features a news article on a new system at Shrewsbury, to help manage business intel - the idea is that the system helps address the lack of capacity for number crunching and analysis.
http://www.e-health-insider.com/news/5900/shrewsbury_pilots_neutrino_bi

"The software displays four bars, including how many people have breached the 18 week promise, those patients that are yet to be see, whether the data was right and can be validated, and a backlog of patients."