Discussion

In England and Northern Ireland in 2009 and 2010, there were more episodes of cataract surgery in children aged under 2 years reported through active surveillance in the IOLu2 study than through routine returns to the NHS Information Centres. The episodes of surgery identified through the BCCIG were limited to children diagnosed with congenital and infantile cataract, while episodes of lens extraction for any pathology were reported to the NHS centres. Lens extraction may be undertaken for reasons other than congenital and infantile cataract, for example, for cases of traumatic cataract, cataract due to previous intraocular surgery, and non-cataractous lens disease such as ectopia lentis. The prevalence or incidence of these other lens-related diseases in children under 2 years old is unknown, but they account for approximately 10% of all children aged under 16 years undergoing lens surgery.7 Thus, the NHS centres should have identified up to 10% more episodes of lens extraction than were reported through the active surveillance network of clinicians identifying cases of congenital/infantile cataract extraction.

The under-reporting by hospitals of paediatric cataract surgery that we have found is consistent with previous reports of poor levels of completeness of centralised NHS ‘returned’ data for other specialities such as paediatric cardiothoracic surgery, adult oncology and psychiatry.8–10 Summary measures derived from the central database statistics are used in a variety of administrative and financial pathways, including Government Treasury negotiations for NHS funds, the allocation of funds within the NHS and the identification of trends in specific conditions or performance of operative procedures. The latest Department of Health audit into NHS clinical coding accuracy identified an overall clinical coding error of 10%, with poor clinical documentation, inadequate training of hospital coders and inadequate involvement of clinical staff in the coding process contributing to inaccurate or incomplete data entry.11 This study is unable to quantify the clinical coding error with regards to the method of cataract extraction in children aged under 2 years, as data on the type of surgery undertaken is only available for a proportion of children identified by the BCCIG. The majority of cataract extraction procedures undertaken in children under 2 years in the UK and reported to the databases were coded under the ‘other extraction of lens’ (lensectomy) category. Lensectomy for early childhood cataract, as first popularised in the UK in the early 1980s following the growing popularity of the use of the vitrectomy handpiece in ophthalmology, is defined as removal of the whole lens complete with anterior and posterior capsules and the anterior vitreous.7 However, as modern paediatric cataract surgery involves consideration of either primary or secondary IOL implantation,1 a larger peripheral capsular ring is left behind to provide intracapsular or sulcus support. This procedure can thus more correctly be defined as an extracapsular lens extraction, which may then be followed by a posterior capsulotomy and anterior vitrectomy. Thus, dependent on the surgical definition used, either all 373 of the operated eyes reported by the BCCIG underwent ‘extracapsular lens extraction’, or 320 of the eyes reported by the BCCIG underwent ‘adapted lensectomy’ with 53 undergoing ‘extracapsular lens extraction’. In either case, there was inaccurate and under-reporting of surgical procedures to the NHS databases. The inaccurate reporting of activity to central NHS databases is partly attributable to the use of multiple codes across different hospitals within each country for paediatric cataract surgery, and to the failure by the international ophthalmic community to update the terminology used in modern paediatric cataract surgery.

The NHS is currently facing significant re-organisation and financial constraints. Accurate coding of hospital activity is vital for the commissioning and payment for activities undertaken and for efficient planning of hospital services. The findings presented here suggest that NHS ophthalmic departments may be disadvantaged by failing to ensure that their administrative system accurately records their true level of activity. Clinical governance is part of the duties of every doctor, and the ophthalmic community should play a part in ensuring that the hospital administrative staff responsible for reporting ophthalmic activity to the national databases have the knowledge and understanding they need to generate accurate and complete reports.

Improvements in routine data collection should allow the HAS system to be used for monitoring trends in the NHS. However, the active surveillance schemes established in the UK that have been successful in the paediatric and ophthalmic fields (the British Paediatric and British Ophthalmic Surveillance Units, BPSU and BOSU), and the high level of ascertainment achieved by the BCCIG in the IOLu2 study highlight the strength of active surveillance and a multicentre collaborative approach in research in uncommon diseases.