Guest Blog: National Statistics – The Road to Accreditation

By way of background, Information Unit had always produced a wealth of information across primary care: medical, ophthalmic, dental and pharmaceutical services for internal use by our health policy and operational colleagues. We even, on occasion, publicly released various ad-hoc reports and tables but were a long way short of being a Code-compliant organisation. This despite BSO being a legally specified producer of Official Statistics (OS) since 2012. It’s not that we didn’t recognise the importance of compliance, it was simply a matter of resources and carving out sufficient time from our day-to-day analytical project support, to progress our OS aspirations. The key was good planning and accepting that this would not be done overnight, taking whatever time was necessary to update our processes, plug identified gaps and develop documentation. Of course, we needed the backing of senior management if we were going to be “distracted” from the day job so we first had to sell the benefits. 

Our first milestone was to release an Annual Compendium in 2018, covering all of our key primary care areas, complying with the Code as far as possible. A quarterly series soon followed and by our third year, responding to user demand, we were ready to split the compendium into separate service areas with a dedicated team responsible for each. User engagement was a key component of the work programme with readership surveys, supplemented by targeted stakeholder interviews, allowing each release to evolve in a way most beneficial to its user base. The fact that our teams were both users, through our ongoing project work, and producers of the data helped enormously in improving its quality and offering guidance on its use. 

By our fourth year, and following two successive releases of our service specific publications, we were finally ready to push the button and subject our outputs to the all seeing eye of the OSR. Yes, it had taken a few years to get to this point, albeit from a fairly low base, but we now felt confident that our processes were in order and we had a good story to tell. The invite was duly issued in November ‘21 and we had our assessment initiation meeting with the OSR team, headed by Dr Anna Price, in January this year. Everything was clearly explained to us in terms of how the process would run and what would be expected from us by way of evidence. The OSR team, comprising Anna, Jo Mulligan and Sarah Whitehead, were really open and friendly (surely a ruse to get us to lower our guard lol) and keen to help with various initiatives that we were planning such as the introduction of Reproducible Analytical Pipelines into our production process. It all seemed reasonably straightforward and, certainly as a veteran of 4 previous assessment campaigns in other NI Departments, much less formal and bureaucratic than I had remembered – can’t last, I thought! 

Roll forward to February, and we had our follow up meeting with the assessment team to discuss our submitted evidence but also, importantly, draw upon information the team had gleaned from our users. This meeting involved all of our publication leads so, with virtual flak jackets donned, we braced ourselves for the inevitable onslaught. But, again, we were pleasantly surprised. The meeting was more like an interesting chat around our various processes, with helpful suggestions and resources offered which could further enhance our outputs. Of course, there were queries and clarifications sought, some of which were followed up in writing in the weeks that followed, but these were conveyed in a constructive way and the different perspective offered us an opportunity to highlight aspects of our process that we’d overlooked in our initial evidence submission.

We received first sight of our draft assessment report the following month and I admit to opening the document with a feeling of slight trepidation. I’d had the impression that the team felt we were in reasonable shape from our meetings but there’s always requirements, I mean they have to find something, right? But no, I read the report twice, definitely no requirements! What it did contain was a succinct summary of how we matched up against the Code of Practice pillars of Trustworthiness, Quality and Value along with some helpful suggestions of where we could enhance our outputs against these. Several useful resources which we might find useful in this regard were also signposted. We had an opportunity to suggest any amendments and correct any factual inaccuracies, of which there were very few, and several weeks later we were notified that the report and recommendation to designate our outputs as National Statistics had been accepted by OSR Regulation Committee. Our journey was finally complete! Of course, it never really ends and we will need to continue to improve and innovate to ensure standards are maintained and the needs of ever more demanding users continue to be satisfied.

Before signing off, I thought I’d leave you with what I feel were the three most important factors in helping us achieve our designation so painlessly.

• Invest time achieving proper buy-in from senior management within your organisation – you will need their support to allow you to spend time developing aspects of your processes that they may not immediately see as being important to their core business.
• Prepare, Prepare, Prepare – Don’t rush to get your outputs assessed, wait until you are properly ready. We were also able to draw upon the support of our colleagues in the Northern Ireland Statistics and Research Agency (NISRA) who provided lots of good advice and resources, including other relevant assessment reports. The assessment focus can change over time and statisticians are constantly innovating. We can learn a lot from our peers. I’m not saying take 4 years but, if you invite an assessment too early, then you will leave yourself with a limited window, typically 3 months, in which to meet requirements. This could feel like a burden on top of your business as usual. Better to meet as many potential requirements ahead of time as possible.
• Have similar outputs assessed as a batch – it might seem tempting to submit individual outputs for assessment in order to make the process more manageable or you may feel that some are more ready than others. However, there can be synergies between outputs and processes that make sense to consider together. We also included all of our publication leads when we met with the assessment team and this all helped deliver a more rounded and efficient assessment.

In my experience, the assessment process itself has definitely evolved for the better and feels more like a collaborative venture these days rather than a statistical audit. It definitely feels more light touch than previously and, although a lot of hard yards are still required to ready your outputs, it is great to see that your efforts will be recognised by the assessment team. 

The Family Practitioner Services statistics in Northern Ireland assessment report has been published today, less than six months since the initial invite for assessment. 

All-in-all we found it to be a very worthwhile and positive experience so if you are thinking of taking the plunge then go for it, you might just be pleasantly surprised!