Statins for primary prevention in practice: clinical use in Wales and the NICE guideline effect

Cardiovascular disease is the biggest killer in the UK, with mortality rates being above average in Wales 1. Statins are used for, and have a proven benefit in, the prevention of cardiovascular disease 2 but are known to have variable patient compliance in clinical trials, with limited data on typical clinical use 3.

There is wide variation in the comparative cost‐effectiveness of different statins within different groups of patients. This was made more complex by a multitude of large clinical trials investigating different groups, statins and doses. This resulted in heterogeneous prescribing practices across the UK, leading to potential reduced cost‐effectiveness at a population level. In 2008, the National Institute for Health and Care Excellence (NICE) published guidance for statin prescription in the UK to maximize the cost–benefit ratio 4.

To investigate the typical clinical (‘real‐life’) use of statins for primary prevention of cardiovascular disease in Wales, a retrospective longitudinal study was performed using the Secure Anonymised Information Linkage (SAIL) databank – a national Welsh database that links person‐based data from primary and secondary healthcare records using robust anonymization techniques 5. A total of 152 781 patients who were prescribed a first statin for the primary prevention of cardiovascular disease between 2000 and 2014 were identified from general practice records by excluding all secondary prevention indications. This group was followed up for five years to examine the choice of statins prescribed, adherence rates in typical clinical practice and the effect of the NICE guidelines (CG67, 2008) on prescribing patterns 4.

Simvastatin was the most commonly initiated statin (74.8%) for the primary prevention of cardiovascular disease in Wales, with most patients remaining on their initial statin type at one‐, three‐ and five‐year follow‐up. The discontinuation rates of all statins were 26.8%, 32.6% and 36.0% at one‐, three‐ and five‐year follow up, respectively. Following the introduction of the NICE CG67 guidelines in 2008, 85.5% of participants were initiated on guideline‐recommended therapy with low‐acquisition‐cost statins at the time (simvastatin 40 mg/20 mg, pravastatin 40 mg; see Table 1).

Table 1.

Initial choice of statins for primary prevention in Wales

Type of statin and dose	Total		Pre‐May 2008		May 2008 onwards
	n	%	n	%	n	%
Atorvastatin	26 935	17.6	22 000	24.1	4935	8.0
10 mg	17 808	66.1	15 697	71.4	2111	42.8
20 mg	7049	26.2	5210	23.7	1839	37.3
40 mg	1847	6.9	1020	4.6	827	16.8
80 mg	231	0.9	73	0.3	158	3.2
Fluvastatin	2033	1.3	1999	2.2	34	0.1
20 mg	1215	59.8	1200	60.0	15	44.1
40 mg	753	37.0	744	37.2	9	26.5
80 mg	65	3.2	55	2.8	10	29.4
Pravastatin	5890	3.9	5222	5.7	668	1.1
10 mg	2531	43.0	2352	45.0	179	26.8
20 mg	2047	34.8	1771	33.9	276	41.3
40 mg	1312	22.3	1099	21.0	213	31.9
Rosuvastatin	3672	2.4	3066	3.4	606	1.0
5 mg	232	6.3	71	2.3	161	26.6
10 mg	3214	87.5	2869	93.6	345	56.9
20 mg	205	5.6	117	3.8	88	14.5
40 mg	21	0.6	9	0.3	12	2.0
Simvastatin	114 252	74.8	59 108	64.7	55 144	89.8
10 mg	16 588	14.5	13 801	23.3	2787	5.1
20 mg	46 342	40.6	29 029	49.1	17 313	31.4
40 mg	51 156	44.8	16 187	27.4	34 969	63.4
80 mg	166	0.1	91	0.2	75	0.1

The present study revealed that in the last 15 years, simvastatin has been the most commonly initiated statin in the primary prevention of cardiovascular disease, although prior to the NICE guidelines there had been no specific pattern in the use of statins. The percentage of patients initiated on low‐acquisition‐cost statins was 50.5% compared with a majority of 85.5% following the publication of the guidelines. Atorvastatin, which was far more expensive at the time, was the second most commonly initiated statin, at 24.1% prior to the implementation of the guidelines. The choice of statin dose varied (Table 1). Indeed, a large proportion of patients were on doses that were lower than recommended by the guidelines and than clinical trial‐proven doses. It is difficult to pinpoint what led to this but it is likely that patient factors (e.g. prejudice by the media) and prescriber factors (e.g. minimizing side effects) have influenced the choice of statin and dose.

The cost implications of statins on health resources in Wales are significant. Atorvastatin came off patent in 2011; prior to that, the cost of a 28‐tablet pack of atorvastatin cost in excess of £25, ten times more than it cost after 2011 6. Taking this into consideration, there would have been a potential saving within NHS Wales of £550 000 per month or £6.6 million per annum had all 22 000 patients who were started on atorvastatin for primary prevention prior to the implementation of the guidelines used simvastatin instead.

Indeed, a number of studies have suggested that physicians may be unaware of the costs of many commonly used drugs, and it was found to be common to underestimate the costs of expensive drugs, and vice versa 7. To address this, NHS Wales publishes monthly prescription cost analyses that reveal the actual cost of all medications dispensed in Wales. In addition, all health boards across Wales utilize ScriptSwitch (Optum, London), a digital aid that advises general practitioners of cost‐effective alternatives at the time of prescribing, in order to promote national guidelines and cost information. However, further interventions to encourage the use of these resources by physicians need to be considered. Nevertheless, the current study suggested that the NICE guidelines have played a pivotal role in resource management in Wales.

Discontinuation rates were higher than the previously described estimates published by NICE 8 of 13% at three years for primary prevention. It was not possible to ascertain the reasons for discontinuation of statins using our current methodology. This is because discontinuation reasons are not routinely coded into primary and secondary care patient records using standardized codes. Instead, they tend to be entered in free‐text format, which makes its upload into SAIL and subsequent extraction for analysis difficult to achieve. One way to improve this is by encouraging the use of the standardized 10th revision of the International Statistical Classification of Diseases and Related Health Problems (ICD‐10; e.g. Z91.1 – noncompliance with medication) or READ codes (e.g. U60CA – adverse reaction to statin). Scandinavian countries have improved the accuracy of coding by linking it to outcome‐based payment schemes 9.

The main strengths of the present study were the large number of typical clinical practice participants in the analysis, which provided a robust and comprehensive analysis of a representative sample of the population in Wales over a number of years. Furthermore, the typical clinical practice nature of this sample population limited the possibilities for selection bias, which may be an issue with carefully selected clinical trial sample populations, and included factors that affect the typical clinical effectiveness of statin therapy which may not have been experienced by trial participants.

Finally, the present study demonstrated the feasibility of using and linking anonymized, routinely collected data in the form of the SAIL databank to avoid bias, protect patient confidentiality and include all patients, thus minimizing attrition and greatly reducing costs. Indeed, clinicians and policy makers may view it as a robust method for national therapeutic audits and quality improvement across Wales.

Competing Interests

All authors have completed the Unified Competing Interest form at www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) and declare: no support from any organization for the submitted work; no financial relationships with any organizations that might have an interest in the submitted work in the previous 3 years; JH declares receiving a one‐off consultancy fee from Optum in May 2015 for delivering a communications workshop.

Protty, M. B. , Lacey, A. , Hayes, J. , and Freeman, P. (2016) Statins for primary prevention in practice: clinical use in Wales and the NICE guideline effect. Br J Clin Pharmacol, 82: 892–894. doi: 10.1111/bcp.13014.