Correspondence

and prevent deaths. The casemix standardised 30-day mortality odds ratio for UK versus Sweden was 1·37.1 The differences occur in the hospitals before discharge, as the mortality rate from discharge up to day 30 is similar between the two countries. However, less than half of the mortality odds ratio difference (1·37 to 1·21) is explained by in-hospital treatments. The main clinical recommendation is to increase the number of primary percutaneous coronary intervention (PCI) in the UK, which has already happened after the implementation of a national policy for primary PCI in 2008. The authors suggest that mortality differences might be even wider, as patients in the UK not captured in the register are older. However, a study2 showed that the coverage for the Swedish registry is about 50% and the mortality rate is two-fold higher for the patients not included in the registry. Hence, despite a study population of more than half a million patients and a casemix model including 17 variables, Chung and colleagues’ study does not guide clinicians on what to improve. Unadjusted selection bias could also decrease the observed differences. The annual report of the Swedish registry states that “it cannot be over-emphasised that…data need to be interpreted with a high degree of caution”.3 We agree. We declare no competing interests.

*Johan Ejerhed, Helena Nordenstedt, Ann-Charlotte Laska [email protected] Danderyd Hospital, Division of Internal Medicine, 182 88 Stockholm, Sweden (JE, HN, A-CL); and Department of Public Health Sciences, Karolinska Institutet, Stockholm, Sweden (HN) 1

2

3

304

Chung S-C, Gedeborg R, Nicholas O, et al. Acute myocardial infarction: a comparison of shortterm survival in national outcome registries in Sweden and the UK. Lancet 2014; 383: 1305–12. Aspberg S, Stenestrand U, Köster M, Kahan T. Large differences between patients with acute myocardial infarction included in two Swedish health registers. Scand J Public Health 2013; 41: 637–43. SWEDEHEART. Annual report 2012. http://www. ucr.uu.se/swedeheart/index.php/arsrapporter/ doc_download/254-swedeheart-annual-report2012-english (accessed Jan 31, 2014).

Sheng-Chia Chung and colleagues1 suggest that the lower 30-day mortality rate after myocardial infarction in Sweden compared with the UK might be due to differences in patient care. They provide a compelling argument for the need of additional comparisons with data from a US population registry indicating a steady decline in ST-segment elevation myocardial infarction (STEMI, 47% in 1999 vs 23% in 2008).2 However, the prevalence of STEMI in the Swedish (32% in 2004, and in 2010) and UK (37% in 2004, and 39% in 2010) registries did not decline; rather there is a precipitous drop in maximum cardiac troponin concentrations during this timeframe in the non-STEMI patients. Specifically, for troponin T the median peak concentration in non-STEMI patients decreased from 0·58 and 0·56 μg/L (in 2004) to 0·378 and 0·33 μg/L (in 2010) in Sweden and the UK. Unfortunately, comparison between mortality rates from 2004 to 2010 within each registry for the non-STEMI group with troponin T was not done. For cardiac troponin I concentration, there is a decrease in non-STEMI patients; however, because these assays are not standardised it might not be appropriate to combine cardiac troponin I concentrations; rather it is recommended to report results as multiples of the 99th percentile.3 For non-STEMI patients analyses are needed to assess if the decrease in cardiac troponin concentration from 2004 to 2010 have translated to a decrease in mortality. This would support the observation for non-STEMI patients that uptake of guideline-directed therapies has decreased the extent of injury and improved outcomes. I received grants and honoraria from Abbott Laboratories, Beckman Coulter, Randox Laboratories, and Roche Diagnostics; and I am a listed inventor on patents filed by McMaster University related to laboratory testing in acute cardiac care.

Peter A Kavsak [email protected] Juravinski Hospital and Cancer Centre, Hamilton, ON L8V 1C3 Canada

1

2

3

Chung S-C, Gedeborg R, Nicholas O, et al. Acute myocardial infarction: a comparison of shortterm survival in national outcome registries in Sweden and the UK. Lancet 2014; 383: 1305–12. Yeh RW, Sidney S, Chandra M, et al. Population trends in the incidence and outcomes of acute myocardial infarction. N Engl J Med 2010; 362: 2155–65. Thygesen K, Alpert JS, Jaffe AS, et al. Third universal definition of myocardial infarction. Circulation 2012; 126: 2020–35.

Does a greater degree of socioeconomic inequality explain worse short-term prognosis after acute myocardial infarction in the UK compared with Sweden? Sheng-Chia Chung and colleagues1 report that short-term outcomes in acute myocardial infaction are worse in the UK compared with Sweden. An apparent difference between the countries that is known to affect clinical outcome and that could explain their findings is the degree of socioeconomic inequality.2 Chung and colleagues used nationwide registries to assess 510 863 patient records, 50 682 of whom died within 30 days. Although data were missing for important variables in a substantial proportion of patients and the majority of patients and events were derived from the UK, such a large number of events lends great power to their analysis. 30 day mortality was 37% higher in the UK (95% CI 1·30–1·45) than in Sweden. Hence a clinically relevant difference between these countries in prognosis after acute myocardial infarction is apparent. Although the authors attempt to explain this difference by higher rates of primary percutaneous coronary intervention and more frequent prescription of β blockers in Sweden, the standardised mortality ratio was estimated to decrease only modestly, from 1·37 to 1·31, if these treatments had been the same in the two countries. Furthermore, adherence to other evidence-based secondary prevention therapy was greater in the UK.1 Hence, differences in therapy are unlikely to explain the difference in mortality. Similarly, no other covariates included in their analyses appear www.thelancet.com Vol 384 July 26, 2014

Correspondence

We declare no competing interests.

*Björn Redfors, Yangzhen Shao, Elmir Omerovic [email protected] Department of Molecular and Clinical Medicine, Institute of Medicine, Gothenburg, Sweden (BR, YS, EO); and Department of Cardiology, Sahlgrenska University Hospital, Gothenburg, Sweden (BR) 1

2

3

4

Chung S-C, Gedeborg R, Nicholas O, et al. Acute myocardial infarction: a comparison of shortterm survival in national outcome registries in Sweden and the UK. Lancet 2014; 383: 1305–12. Mackenbach JP, Cavelaars AE, Kunst AE, Groenhof F. Socioeconomic inequalities in cardiovascular disease mortality; an international study. Eur Heart J 2000; 21: 1141–51. Marmot M, Allen J, Bell R, Bloomer E, Goldblatt P. WHO European review of social determinants of health and the health divide. Lancet 2012; 380: 1011–29. Hales S, Howden-Chapman P, Salmond C, Woodward A, Mackenbach J. National infant mortality rates in relation to gross national product and distribution of income. Lancet 1999; 354: 2047.

that lives could have been saved if the Swedish pattern of care had been replicated in the UK is misguided. The apparent difference in 30-day mortality between the UK and Sweden might be due to case ascertainment rates,2,3 or data collection practices.3,4 The media coverage5 that followed the publication of the Article1 attributed apparent excess deaths to deficiencies in NHS care and was unfair. As academics and members of the Steering Group of the Myocardial Ischaemia National Audit Project (MINAP)—the source of the UK data for that Article1—we feel a duty to bring our concerns about the interpretation of this work to the attention of The Lancet readers. We have published additional detail elsewhere.6 The views expressed are the authors’ and not those of their respective institutions. ENL, LG, and CW are employed by University College London (UCL) and work within the National Institute for Cardiovascular Outcomes Research (NICOR). ENL, LG, TQ, and CW are members of the Steering Group of the Myocardial Ischaemia National Audit Project at NICOR. Some of the authors of the paper1 on which we comment are also UCL employees and members of NICOR, and will appraise ENL, LG, and CW.

*Emmanuel N Lazaridis, Lucia Gavalova, Simon Jones, Tom Quinn, Clive Weston [email protected] University College London, London WC1E 6BT, UK (ENL, LG); University of Surrey, Guildford, UK (SJ, TQ); and Swansea University, Swansea, UK (CW) 1

Sheng-Chia Chung and colleagues1 compared 30-day mortality following admission with acute coronary syndrome to hospitals in either Sweden or England and Wales. We do not agree with the interpretation of the analyses. Although the authors explain the difference in mortality rates between the two national registries in terms of casemix and treatment factors, the confidence intervals around their estimates indicate no evidence that this difference arises from differential percutaneous coronary intervention and β blockers use, and that most of the difference cannot be explained by the combination of treatment factors considered. Thus, the message www.thelancet.com Vol 384 July 26, 2014

2

3

4

5

6

Chung S-C, Gedeborg R, Nicholas O, et al. Acute myocardial infarction: a comparison of shortterm survival in national outcome registries in Sweden and the UK. Lancet 2014; 383: 1305–12. Herrett, E, Shah AD, Boggon R, et al. Completeness and diagnostic validity of recording acute myocardial infarction events in primary care, hospital care, disease registry, and national mortality records: cohort study. BMJ 2013; 346: f2350. SWEDEHEART. Annual report 2012. http://www. ucr.uu.se/swedeheart/index.php/arsrapporter/ doc_download/254-swedeheart-annual-report2012-english (accessed July 14, 2014). Gavalova L, Weston C, Birkhead J, et al. Myocardial Ischaemia National Audit Project (MINAP): How the NHS cares for patients with heart attack. Tenth Public Report 2011. http://www.hqip.org. uk/assets/NCAPOP-Library/MINAP-publicreport-2011.pdf (accessed July 14, 2014). BBC. UK heart-attack survival rate ‘should have been better’. 23 January 2014. http://www. bbc.co.uk/news/health-25841930 (accessed July 14, 2014). Lazaridis EN, Gavalova L, Jones S, et al. The UK vs Sweden: Is the NHS really so bad? PeerJ PrePrints 2014; 2: e253v1.

Authors’reply We thank the correspondents for their comments on our Article. 1 We agree with Johan Ejerhed and colleagues that understanding selection biases is vital for interpretation and we have previously shown that MINAP, like SWEDEHEART, does not capture all myocardial infarctions identified in hospital discharge data or in primary care. 2 These missing patients are likely to be sicker; it is plausible that a higher proportion of patients are missed in the UK than in Sweden and this could mean that the true mortality differences between countries are even greater. Further research is required using multiple national sources of data including primary care.3 Our paper1 does offer a guide to the clinician—clinicians have a responsibility to improve the quality, completeness, and comparability of clinical data and to address the slow diffusion of new technologies in the UK. Peter Kavsak suggests widening comparisons to include US populations; we have begun this and noted surprising international differences in clinical management of non ST-segment elevation myocardial infarction (NSTEMI). 4 We have done new analyses and can confirm that the decreases in troponin in patients with NSTEMI are associated with lower mortality (figure, appendix). The decreasing troponin is consistent with improvements in care leading to smaller infarcts and better outcomes. We agree that socioeconomic status is associated with risk of acute myocardial infarction, 5 as Björn Redfors and colleagues suggest. However, wider income inequalities in the UK compared with Sweden are unlikely to explain our results because such inequalities have remained fairly constant during most of the study period,6 and also because our analysis of CALIBER data2 (based on 16 239 patients with myocardial

James Cavallini/Science Photo Library

to explain the observed difference in short-term outcome after acute myocardial infarction.1 Previous studies2–4 have established a relation between health outcome (including cardiovascular disease) and socioeconomic inequality. Because socioeconomic inequalities are greater in the UK than in Sweden, they could explain Chung and colleagues’ findings. 1 Future studies should include socioeconomic variables in their statistical models.

See Online for appendix

305

International comparisons of acute myocardial infarction.

International comparisons of acute myocardial infarction. - PDF Download Free
107KB Sizes 1 Downloads 3 Views