Correspondence
Hospitalisation for children with rickets in England: a historical perspective Nutritional rickets caused by vitamin D deficiency is associated with a diet deficient in vitamin D, inadequate exposure to sunlight, and reduced endogenous synthesis of vitamin D in the skin. The Chief Medical Officer for England has recently drawn attention, in her annual report, to the continuing problem of vitamin D deficiency in children.1,2 We have analysed rates of hospital admissions for rickets in England, across five decades, in children younger than 15 years. We analysed data for all England using routinely collected hospital episode statistics for 1968– 2011 (no national data for 1986–90). We also analysed data for the former Oxford NHS Region (population about 2·5 million) between 1963 and 2011 using the Oxford record linkage study (ORLS),3 which includes data for all day cases and inpatients. Data in the ORLS from 1963 and in England from 1999 were linked for each 7 6
individual to distinguish episodes of care (individuals can have more than one episode of care, episode-based analyses) and individuals receiving care (irrespective of the number of episodes, individual-based analyses). Analyses were done by selecting records with International Classification of Diseases (ICD) codes for active rickets (ICD7 code 283, ICD8 265·0, ICD9 268·0, and ICD10 E55·0). Rates were directly standardised using 5-year age groups and the European standard population. Annual rates—episode-based and individual-based—are shown in the figure. Hospitalisation rates for rickets were low in the 1960s and 1970s and declined further in the 1980s and 1990s. They increased in the 2000s. In ORLS, rates were lowest in 1991–96 at 0·34 children per 100 000 under 15 years (95% CI 0·13–0·55) and highest in 2007–11 at 1·78 (1·27–2·29). In all England in 2007–11, the episode-based rate was 4·78 (4·58–4·99) and the personbased rate was 3·16 (3·00–3·33) per 100 000 aged under 15. Most patients were younger than 5 years. Ethnic origin is not well recorded
England: episode-based rate England: individual-based rate Oxford: individual based rate
Rate per 100 000
5 4
in Hospital Episode Statistics but, where recorded, 32% of the people with rickets were white (compared with 83% in the general population younger than 15 years),4 33% south Asian (vs 8%), 33% black (vs 3%), and 2% other non-white (vs 6%). Their duration of residence in the UK is unknown. Between 2001 and 2009, the number of white children in England fell by 6% and the number of non-white children increased by 19%.4 The increase in rates might be a reflection of increased incidence or severity of rickets, or changes in admission thresholds, diagnostic, or coding practices. However, with other reports of a recent increase in rickets,2,5 the previous hypothesis was that a rise in hospitalisation rates might occur. The coding of rickets is straightforward, given the diagnosis; and the same exact term, active rickets, was used in the ICD in all decades studied. Between 1968 and 1985, Hospital In-Patient Enquiry and ORLS were collected independently of each other; each shows the same profile suggesting face validity at least at that time. Hospitalisation rates for rickets in England are now the highest in five decades. Currently, rickets is not confined to non-white people, though it is more common in non-white than white individuals. There is a case for a national confidential audit of rickets. We declare that we have no conflicts of interest.
3
*Michael Goldacre, Nick Hall, David G R Yeates
2
[email protected] 1
Unit of Health-Care Epidemiology, Nuffield Department of Population Health, University of Oxford, Oxford OX3 7LF, UK
19 63 19 65 19 6 19 7 69 19 71 19 73 19 75 19 77 19 79 19 81 19 83 19 85 19 8 19 7 89 19 91 19 93 19 95 19 9 19 7 99 20 01 20 03 20 05 20 0 20 7 09 20 11
0
1
Year
Figure: Rickets in children younger than 15 years: age-standardised rates per 100 000 younger than 15 years in England and in Oxford Red: Episode-based rates (hospital discharges) for England, 1968–2011. Between 1968 and 1985, the data were a one in 10 sample of all NHS hospital discharges (excluding day cases), then termed the Hospital In-Patient Enquiry, and we scaled to 100% by multiplying each number of annual admissions for rickets by 10. There were no national data for 1986–90. Between 1990 and 2011, the Hospital Episode Statistics included day cases (patients admitted to hospital who did not stay overnight) and inpatients. Green: individual-based rates for England, from 1999. Blue: individual-based rates for Oxford, from the Oxford record linkage study.3
www.thelancet.com Vol 383 February 15, 2014
2
3
Davies SC, Lemer C, Strelitz J, Weil L. Our children deserve better: prevention pays. Lancet 2013, 382: 1383–84. Davies SC. Annual report of the Chief Medical Officer. Our children deserve better: prevention pays. London: Department of Health, 2012. Chou MR, Malik AN, Suleman M, Gray M, Yeates D, Goldacre MJ. Time trends over five decades, and recent geographical variation, in rates of childhood squint surgery in England. Br J Ophthalmol 2013; 97: 746–51.
Submissions should be made via our electronic submission system at http://ees.elsevier.com/ thelancet/
597
Correspondence
4
5
Office for National Statistics. Population estimates by ethnic group. http://www.ons. gov.uk/ons/taxonomy/index.html?nscl= Population+Estimates+by+Ethnic+Group#tabdata-tables (accessed Dec 23, 2013). Ahmed SF, Franey C, McDevitt H, et al. Recent trends and clinical features of childhood vitamin D deficiency presenting to a children’s hospital in Glasgow. Arch Dis Child 2011; 96: 694–96.
Refining the American guidelines for prevention of cardiovascular disease We would like to propose a compromise in the debate on the clinical application of the recently proposed American guidelines for prevention of cardiovascular disease.1 Paul Ridker and Nancy Cook (Nov 30, p 1762)2 criticise the current guidelines’ assumption of a constant relative risk reduction and its subsequent focus on absolute (baseline) risk predictions. They plea for statins prescription based on treatment effects observed in specific trial populations. In our view, combining absolute risk predictions with individualised estimates of relative risk reduction is required to quantify the absolute treatment benefit of statins. Relative risk reduction across subgroups with different levels of baseline risk can be based on the results of the individual patient data meta-analysis of statin trials. 3 Ideally, the individualised relative risk reductions are estimated in a re-analysis of these trial data, by adding a statistical treatment interaction with the risk predictions according to 2013 guidelines.4 Ridker and Cook recommend recalibration of the guidelines’ new prediction model in additional external validation cohorts. Rather, we should use already available contemporary validation cohorts to adjust poorly calibrated risk predictions for time trends. To account for well recognised cardiovascular disease risk differences across the ethnic groups of Hispanics, Asians, and native-Americans, 598
recalibration might be based on available external data as well.5 In conclusion, we recommend building guidelines on adequate estimates of absolute treatment benefit, requiring a recalibrated absolute risk prediction model in conjunction with individualised estimates of the relative risk reduction. We declare that we have no conflicts of interest.
*David van Klaveren, Yvonne Vergouwe, Ewout W Steyerberg [email protected] Department of Public Health, Erasmus MC, 3015 Rotterdam, The Netherlands 1
2
3
4
5
Stone NJ, Robinson J, Lichtenstein AH, et al. 2013 ACC/AHA guideline on the treatment of blood cholesterol to reduce atherosclerotic cardiovascular risk in adults: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. Circulation 2013; published online Nov 13. DOI:10.1161/01.cir.0000437738.63853.7a. Ridker PM, Cook NR. Statins: new American guidelines for prevention of cardiovascular disease. Lancet 2013; 382: 1762–65. Cholesterol Treatment Trialists’ (CTT) Collaborators. The effects of lowering LDL cholesterol with statin therapy in people at low risk of vascular disease: meta-analysis of individual data from 27 randomised trials. Lancet 2012; 380: 581–90. Kent DM, Rothwell PM, Ioannidis JP, Altman DG, Hayward RA. Assessing and reporting heterogeneity in treatment effects in clinical trials: a proposal. Trials 2010; 11: 85. D’Agostino RB Sr, Grundy S, Sullivan LM, Wilson P, Group CHDRP. Validation of the Framingham coronary heart disease prediction scores: results of a multiple ethnic groups investigation. JAMA 2001; 286: 180–87.
The recent American Heart Association (AHA) and American College of Cardiology (ACC) guidelines1 recommend that more people are offered treatment to prevent heart attacks and strokes; they do this by lowering the risk cutoff from a 20% 10-year risk to 7·5% (the approximate risk of a person aged 60 years). Paul Ridker and Nancy Cook state that the risk calculator in the guidelines overestimates risk about two-fold.2 This variation, however, has little effect on discriminating between who will and will not have a heart attack or stroke (ie, on screening performance).3 Age has a much greater discriminatory effect than do risk factors such as
blood pressure and cholesterol despite their aetiological importance.3 Screening using age alone has advantages. It is simple and removes the need for risk estimation, which is only necessary when many screening factors are combined (eg, in antenatal screening for Down’s syndrome, where no single marker dominates over others). Therefore, screening using age alone avoids debate about the accuracy of risk estimation using different algorithms, none of which materially improve screening performance over age alone.4 Also, the focus should move from risk in the absence of intervention to the health benefit of intervention, which is what matters. If people took blood pressure and cholesterol lowering medicines from age 60 years without previous risk factor measurement, one third would benefit and they would, on average, gain 7 years of life without a heart attack or stroke. NW jointly holds European, Canadian, and US patents for a combinaton pill for the prevention of cardiovascular disease. JM declares that he has no conflicts of interest.
*Nicholas Wald, Joan Morris [email protected] Wolfson Institute of Preventive Medicine, London, EC1M 6BQ, UK 1
2
3
4
Stone NJ, Robinson J, Lichtenstein AH, et al. 2013 ACC/AHA guideline on the treatment of blood cholesterol to reduce atherosclerotic cardiovascular risk in adults: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. J Am Coll Cardiol 2013; published online Nov 13. DOI:10.1016/j. jacc.2013.11.002. Ridker PM, Cook NR. Statins: new American guidelines for prevention of cardiovascular disease. Lancet 2013; 382: 1762–65. Wald NJ, Simmonds M, Morris JK. Screening for future cardiovascular disease using age alone compared with multiple risk factors and age. PLoS One 2011; 6: e18742. Simmonds MC, Wald NJ. Risk estimation versus screening performance: a comparison of six risk algorithms for cardiovascular disease. J Med Screen 2012; 19: 201–05.
Paul Ridker and Nancy Cook1 express concern that the new American Heart Association (AHA) and American College of Cardiology (ACC) risk calculator “systematically overestimates” observed risks.2 This www.thelancet.com Vol 383 February 15, 2014
Hospitalisation for children with rickets in England: a historical perspective Nutritional rickets caused by vitamin D deficiency is associated with a diet deficient in vitamin D, inadequate exposure to sunlight, and reduced endogenous synthesis of vitamin D in the skin. The Chief Medical Officer for England has recently drawn attention, in her annual report, to the continuing problem of vitamin D deficiency in children.1,2 We have analysed rates of hospital admissions for rickets in England, across five decades, in children younger than 15 years. We analysed data for all England using routinely collected hospital episode statistics for 1968– 2011 (no national data for 1986–90). We also analysed data for the former Oxford NHS Region (population about 2·5 million) between 1963 and 2011 using the Oxford record linkage study (ORLS),3 which includes data for all day cases and inpatients. Data in the ORLS from 1963 and in England from 1999 were linked for each 7 6
individual to distinguish episodes of care (individuals can have more than one episode of care, episode-based analyses) and individuals receiving care (irrespective of the number of episodes, individual-based analyses). Analyses were done by selecting records with International Classification of Diseases (ICD) codes for active rickets (ICD7 code 283, ICD8 265·0, ICD9 268·0, and ICD10 E55·0). Rates were directly standardised using 5-year age groups and the European standard population. Annual rates—episode-based and individual-based—are shown in the figure. Hospitalisation rates for rickets were low in the 1960s and 1970s and declined further in the 1980s and 1990s. They increased in the 2000s. In ORLS, rates were lowest in 1991–96 at 0·34 children per 100 000 under 15 years (95% CI 0·13–0·55) and highest in 2007–11 at 1·78 (1·27–2·29). In all England in 2007–11, the episode-based rate was 4·78 (4·58–4·99) and the personbased rate was 3·16 (3·00–3·33) per 100 000 aged under 15. Most patients were younger than 5 years. Ethnic origin is not well recorded
England: episode-based rate England: individual-based rate Oxford: individual based rate
Rate per 100 000
5 4
in Hospital Episode Statistics but, where recorded, 32% of the people with rickets were white (compared with 83% in the general population younger than 15 years),4 33% south Asian (vs 8%), 33% black (vs 3%), and 2% other non-white (vs 6%). Their duration of residence in the UK is unknown. Between 2001 and 2009, the number of white children in England fell by 6% and the number of non-white children increased by 19%.4 The increase in rates might be a reflection of increased incidence or severity of rickets, or changes in admission thresholds, diagnostic, or coding practices. However, with other reports of a recent increase in rickets,2,5 the previous hypothesis was that a rise in hospitalisation rates might occur. The coding of rickets is straightforward, given the diagnosis; and the same exact term, active rickets, was used in the ICD in all decades studied. Between 1968 and 1985, Hospital In-Patient Enquiry and ORLS were collected independently of each other; each shows the same profile suggesting face validity at least at that time. Hospitalisation rates for rickets in England are now the highest in five decades. Currently, rickets is not confined to non-white people, though it is more common in non-white than white individuals. There is a case for a national confidential audit of rickets. We declare that we have no conflicts of interest.
3
*Michael Goldacre, Nick Hall, David G R Yeates
2
[email protected] 1
Unit of Health-Care Epidemiology, Nuffield Department of Population Health, University of Oxford, Oxford OX3 7LF, UK
19 63 19 65 19 6 19 7 69 19 71 19 73 19 75 19 77 19 79 19 81 19 83 19 85 19 8 19 7 89 19 91 19 93 19 95 19 9 19 7 99 20 01 20 03 20 05 20 0 20 7 09 20 11
0
1
Year
Figure: Rickets in children younger than 15 years: age-standardised rates per 100 000 younger than 15 years in England and in Oxford Red: Episode-based rates (hospital discharges) for England, 1968–2011. Between 1968 and 1985, the data were a one in 10 sample of all NHS hospital discharges (excluding day cases), then termed the Hospital In-Patient Enquiry, and we scaled to 100% by multiplying each number of annual admissions for rickets by 10. There were no national data for 1986–90. Between 1990 and 2011, the Hospital Episode Statistics included day cases (patients admitted to hospital who did not stay overnight) and inpatients. Green: individual-based rates for England, from 1999. Blue: individual-based rates for Oxford, from the Oxford record linkage study.3
www.thelancet.com Vol 383 February 15, 2014
2
3
Davies SC, Lemer C, Strelitz J, Weil L. Our children deserve better: prevention pays. Lancet 2013, 382: 1383–84. Davies SC. Annual report of the Chief Medical Officer. Our children deserve better: prevention pays. London: Department of Health, 2012. Chou MR, Malik AN, Suleman M, Gray M, Yeates D, Goldacre MJ. Time trends over five decades, and recent geographical variation, in rates of childhood squint surgery in England. Br J Ophthalmol 2013; 97: 746–51.
Submissions should be made via our electronic submission system at http://ees.elsevier.com/ thelancet/
597
Correspondence
4
5
Office for National Statistics. Population estimates by ethnic group. http://www.ons. gov.uk/ons/taxonomy/index.html?nscl= Population+Estimates+by+Ethnic+Group#tabdata-tables (accessed Dec 23, 2013). Ahmed SF, Franey C, McDevitt H, et al. Recent trends and clinical features of childhood vitamin D deficiency presenting to a children’s hospital in Glasgow. Arch Dis Child 2011; 96: 694–96.
Refining the American guidelines for prevention of cardiovascular disease We would like to propose a compromise in the debate on the clinical application of the recently proposed American guidelines for prevention of cardiovascular disease.1 Paul Ridker and Nancy Cook (Nov 30, p 1762)2 criticise the current guidelines’ assumption of a constant relative risk reduction and its subsequent focus on absolute (baseline) risk predictions. They plea for statins prescription based on treatment effects observed in specific trial populations. In our view, combining absolute risk predictions with individualised estimates of relative risk reduction is required to quantify the absolute treatment benefit of statins. Relative risk reduction across subgroups with different levels of baseline risk can be based on the results of the individual patient data meta-analysis of statin trials. 3 Ideally, the individualised relative risk reductions are estimated in a re-analysis of these trial data, by adding a statistical treatment interaction with the risk predictions according to 2013 guidelines.4 Ridker and Cook recommend recalibration of the guidelines’ new prediction model in additional external validation cohorts. Rather, we should use already available contemporary validation cohorts to adjust poorly calibrated risk predictions for time trends. To account for well recognised cardiovascular disease risk differences across the ethnic groups of Hispanics, Asians, and native-Americans, 598
recalibration might be based on available external data as well.5 In conclusion, we recommend building guidelines on adequate estimates of absolute treatment benefit, requiring a recalibrated absolute risk prediction model in conjunction with individualised estimates of the relative risk reduction. We declare that we have no conflicts of interest.
*David van Klaveren, Yvonne Vergouwe, Ewout W Steyerberg [email protected] Department of Public Health, Erasmus MC, 3015 Rotterdam, The Netherlands 1
2
3
4
5
Stone NJ, Robinson J, Lichtenstein AH, et al. 2013 ACC/AHA guideline on the treatment of blood cholesterol to reduce atherosclerotic cardiovascular risk in adults: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. Circulation 2013; published online Nov 13. DOI:10.1161/01.cir.0000437738.63853.7a. Ridker PM, Cook NR. Statins: new American guidelines for prevention of cardiovascular disease. Lancet 2013; 382: 1762–65. Cholesterol Treatment Trialists’ (CTT) Collaborators. The effects of lowering LDL cholesterol with statin therapy in people at low risk of vascular disease: meta-analysis of individual data from 27 randomised trials. Lancet 2012; 380: 581–90. Kent DM, Rothwell PM, Ioannidis JP, Altman DG, Hayward RA. Assessing and reporting heterogeneity in treatment effects in clinical trials: a proposal. Trials 2010; 11: 85. D’Agostino RB Sr, Grundy S, Sullivan LM, Wilson P, Group CHDRP. Validation of the Framingham coronary heart disease prediction scores: results of a multiple ethnic groups investigation. JAMA 2001; 286: 180–87.
The recent American Heart Association (AHA) and American College of Cardiology (ACC) guidelines1 recommend that more people are offered treatment to prevent heart attacks and strokes; they do this by lowering the risk cutoff from a 20% 10-year risk to 7·5% (the approximate risk of a person aged 60 years). Paul Ridker and Nancy Cook state that the risk calculator in the guidelines overestimates risk about two-fold.2 This variation, however, has little effect on discriminating between who will and will not have a heart attack or stroke (ie, on screening performance).3 Age has a much greater discriminatory effect than do risk factors such as
blood pressure and cholesterol despite their aetiological importance.3 Screening using age alone has advantages. It is simple and removes the need for risk estimation, which is only necessary when many screening factors are combined (eg, in antenatal screening for Down’s syndrome, where no single marker dominates over others). Therefore, screening using age alone avoids debate about the accuracy of risk estimation using different algorithms, none of which materially improve screening performance over age alone.4 Also, the focus should move from risk in the absence of intervention to the health benefit of intervention, which is what matters. If people took blood pressure and cholesterol lowering medicines from age 60 years without previous risk factor measurement, one third would benefit and they would, on average, gain 7 years of life without a heart attack or stroke. NW jointly holds European, Canadian, and US patents for a combinaton pill for the prevention of cardiovascular disease. JM declares that he has no conflicts of interest.
*Nicholas Wald, Joan Morris [email protected] Wolfson Institute of Preventive Medicine, London, EC1M 6BQ, UK 1
2
3
4
Stone NJ, Robinson J, Lichtenstein AH, et al. 2013 ACC/AHA guideline on the treatment of blood cholesterol to reduce atherosclerotic cardiovascular risk in adults: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. J Am Coll Cardiol 2013; published online Nov 13. DOI:10.1016/j. jacc.2013.11.002. Ridker PM, Cook NR. Statins: new American guidelines for prevention of cardiovascular disease. Lancet 2013; 382: 1762–65. Wald NJ, Simmonds M, Morris JK. Screening for future cardiovascular disease using age alone compared with multiple risk factors and age. PLoS One 2011; 6: e18742. Simmonds MC, Wald NJ. Risk estimation versus screening performance: a comparison of six risk algorithms for cardiovascular disease. J Med Screen 2012; 19: 201–05.
Paul Ridker and Nancy Cook1 express concern that the new American Heart Association (AHA) and American College of Cardiology (ACC) risk calculator “systematically overestimates” observed risks.2 This www.thelancet.com Vol 383 February 15, 2014
