Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Single‐year data not enough

Single‐year data not enough Letters to the Editor conf irming the method of age adjustment used) and clarif ication (e.g. re errors in the article). While we don’t agree with the authors’ decision to use direct standardisation (an unstable measure when John Glover and Sarah Tennant numbers are relatively small), or that other data were not widely Public Health Information Development Unit, The available (since 1997 the HealthWIZ package has included data University of Adelaide, South Australia up to 1995), of most concern is the use of a single year of data on which to base the analysis. This concern arises because of the The Wilkinson et al. study in the June 2000 issue (Aust N Z possible volatility in deaths data on a yearly basis. Journal of Public Health 2000;226-33) provided an analysis of 1997 deaths data by the Australian Bureau of Statistics’ Statistical In their analysis, the Statistical Division of Pilbara in Western Australia’s far north is shown as having the lowest age-standardised Divisions. The article prompted a response from Booth and Smith all-cause mortality per 1,000 population of any Statistical Division in the October 2000 issue (Aust N Z Journal of Public Health in 1997. The single year rate, of 5.78 deaths per 1,000 population 2000;554-5). Wilkinson et al. responded to the Booth and Smith (calculated by direct standardisation), is lower than the Pilbara letter, providing rebuttal (e.g. re lack of other available data and Rate Figure 1: Age- standardised all causes mortality per 1,000 14 population: Pilbara Statistical Division, Western Australia. 1985 '87 '89 '91 '93 '95 '97 '99 1985 ’87 ’89 ’91 ’93 ’95 ’97 ’99 1985 1985 '87 '89 '91 '93 '95 '97 '9 ’87 ’89 ’91 ’93 ’95 ’97 ’99 1986 ’88 ’90 ’92 ’94 ’96 ’98 1986 ’88 ’90 ’92 ’94 ’96 ’98 1986 '88 '90 '92 '94 '96 '98 1986 '88 '90 '92 '94 '96 '98 Pilbara Western Australia Table 1: Age-standardised all-cause mortality per 1000 population. Study Calculated 1997 1995-97 1996-98 1997-99 Direct standardisation Australia 6.98 6.99 6.93 6.84 Western Australia 6.72 6.79 6.75 6.63 Pilbara: Rate 5.78 7.55 7.55 7.04 % difference from single year rate N/A 30.60 30.60 21.80 Calculated Indirect standardisation Australia 6.98 6.99 6.93 6.84 Western Australia 6.73 6.79 6.75 6.64 Pilbara: Rate 7.73 9.05 8.76 8.27 % difference from single year rate N/A 17.10 13.30 6.90 Source: Data supplied by the Australian Bureau of Statistics. 278 AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH 2002 VOL. 26 NO. 3 1 rate in any other year from 1985 to 1999 (see Figure 1). It is also lower than the Western Australian State rate in any of these years. The rate is based on 95 deaths registered to residents of Pilbara in 1997. Registrations in the years immediately before and after 1997 were 115 (1995 and 1996), 113 (1998) and 108 (1999). Neither variations between years in late registrations nor boundary changes appear to have been the cause of the lower number in 1997. Had the 1997 rate been calculated using indirect standardisation, it would have still been low in relation to adjacent years, but not the lowest over the 1985 to 1999 period, nor lower than the Western Australian State rate. Moreover, had the Pilbara rate been calculated for a combination of years, it would not have been lower than the State rate, more truly reflecting the long-term picture for this area. For example, rates for 1995-97, 1996-98 and 1997-99 are all well above that calculated by Wilkinson et al. for 1997 (see Table 1). It is of note that the percentage difference between the single year and three- year combined rates is always less when the indirect method of age standardisation is used. The data in Table 1 support our concerns as to the use of a single year of data in area analyses where numbers are relatively small and movements from year to year are likely to be important. It is practicable to obtain more than one year’s data for deaths, whether from the Australian Bureau of Statistics (primary data), or HealthWIZ (primary data – which can be exported for analysis – and secondary data – which is quite transparent as to the processes used in the production of rates). The main limitations to the practicability of aggregating data over time are the cost of data and the need to deal with boundary changes. These limitations should not be used as an excuse for a lack of rigour. This is especially so when a knowledge of Australian demography should alert the researcher to be wary of a death rate in Pilbara that was lower than that for the State. 1. The indirect and direct standardised rates were calculated using the Australian population for 1997 (the population used in the Wilkinson et al. study) as the standard in each of the years examined. Local rates were calculated using the ERP for the year in question, or the mid-point where three years were involved. Correspondence to: John Glover, PHIDU, 10 Pulteney Street, Adelaide South Australia 5001. Fax: (08) 8303 6240; e-mail: john.glover@adelaide.edu.au 2002 VOL. 26 NO. 3 AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH 279 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Australian and New Zealand Journal of Public Health Wiley

Single‐year data not enough

Download PDF
2 pages

Loading next page...
 
/lp/wiley/single-year-data-not-enough-2OeOoVYJZj

References (0)

References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.

Publisher
Wiley
Copyright
Copyright © 2002 Wiley Subscription Services
ISSN
1326-0200
eISSN
1753-6405
DOI
10.1111/j.1467-842X.2002.tb00689.x
Publisher site
See Article on Publisher Site

Abstract

Letters to the Editor conf irming the method of age adjustment used) and clarif ication (e.g. re errors in the article). While we don’t agree with the authors’ decision to use direct standardisation (an unstable measure when John Glover and Sarah Tennant numbers are relatively small), or that other data were not widely Public Health Information Development Unit, The available (since 1997 the HealthWIZ package has included data University of Adelaide, South Australia up to 1995), of most concern is the use of a single year of data on which to base the analysis. This concern arises because of the The Wilkinson et al. study in the June 2000 issue (Aust N Z possible volatility in deaths data on a yearly basis. Journal of Public Health 2000;226-33) provided an analysis of 1997 deaths data by the Australian Bureau of Statistics’ Statistical In their analysis, the Statistical Division of Pilbara in Western Australia’s far north is shown as having the lowest age-standardised Divisions. The article prompted a response from Booth and Smith all-cause mortality per 1,000 population of any Statistical Division in the October 2000 issue (Aust N Z Journal of Public Health in 1997. The single year rate, of 5.78 deaths per 1,000 population 2000;554-5). Wilkinson et al. responded to the Booth and Smith (calculated by direct standardisation), is lower than the Pilbara letter, providing rebuttal (e.g. re lack of other available data and Rate Figure 1: Age- standardised all causes mortality per 1,000 14 population: Pilbara Statistical Division, Western Australia. 1985 '87 '89 '91 '93 '95 '97 '99 1985 ’87 ’89 ’91 ’93 ’95 ’97 ’99 1985 1985 '87 '89 '91 '93 '95 '97 '9 ’87 ’89 ’91 ’93 ’95 ’97 ’99 1986 ’88 ’90 ’92 ’94 ’96 ’98 1986 ’88 ’90 ’92 ’94 ’96 ’98 1986 '88 '90 '92 '94 '96 '98 1986 '88 '90 '92 '94 '96 '98 Pilbara Western Australia Table 1: Age-standardised all-cause mortality per 1000 population. Study Calculated 1997 1995-97 1996-98 1997-99 Direct standardisation Australia 6.98 6.99 6.93 6.84 Western Australia 6.72 6.79 6.75 6.63 Pilbara: Rate 5.78 7.55 7.55 7.04 % difference from single year rate N/A 30.60 30.60 21.80 Calculated Indirect standardisation Australia 6.98 6.99 6.93 6.84 Western Australia 6.73 6.79 6.75 6.64 Pilbara: Rate 7.73 9.05 8.76 8.27 % difference from single year rate N/A 17.10 13.30 6.90 Source: Data supplied by the Australian Bureau of Statistics. 278 AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH 2002 VOL. 26 NO. 3 1 rate in any other year from 1985 to 1999 (see Figure 1). It is also lower than the Western Australian State rate in any of these years. The rate is based on 95 deaths registered to residents of Pilbara in 1997. Registrations in the years immediately before and after 1997 were 115 (1995 and 1996), 113 (1998) and 108 (1999). Neither variations between years in late registrations nor boundary changes appear to have been the cause of the lower number in 1997. Had the 1997 rate been calculated using indirect standardisation, it would have still been low in relation to adjacent years, but not the lowest over the 1985 to 1999 period, nor lower than the Western Australian State rate. Moreover, had the Pilbara rate been calculated for a combination of years, it would not have been lower than the State rate, more truly reflecting the long-term picture for this area. For example, rates for 1995-97, 1996-98 and 1997-99 are all well above that calculated by Wilkinson et al. for 1997 (see Table 1). It is of note that the percentage difference between the single year and three- year combined rates is always less when the indirect method of age standardisation is used. The data in Table 1 support our concerns as to the use of a single year of data in area analyses where numbers are relatively small and movements from year to year are likely to be important. It is practicable to obtain more than one year’s data for deaths, whether from the Australian Bureau of Statistics (primary data), or HealthWIZ (primary data – which can be exported for analysis – and secondary data – which is quite transparent as to the processes used in the production of rates). The main limitations to the practicability of aggregating data over time are the cost of data and the need to deal with boundary changes. These limitations should not be used as an excuse for a lack of rigour. This is especially so when a knowledge of Australian demography should alert the researcher to be wary of a death rate in Pilbara that was lower than that for the State. 1. The indirect and direct standardised rates were calculated using the Australian population for 1997 (the population used in the Wilkinson et al. study) as the standard in each of the years examined. Local rates were calculated using the ERP for the year in question, or the mid-point where three years were involved. Correspondence to: John Glover, PHIDU, 10 Pulteney Street, Adelaide South Australia 5001. Fax: (08) 8303 6240; e-mail: john.glover@adelaide.edu.au 2002 VOL. 26 NO. 3 AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH 279

Journal

Australian and New Zealand Journal of Public Health – Wiley

Published: Jan 1, 2002

There are no references for this article.