On this page:

Cancer in Scotland: Sustaining Change: Cancer Incidence Projections for Scotland (2001-2020) - An aid to planning cancer services

« Previous | Contents |

Listen

Cancer in Scotland: Sustaining Change
Cancer Incidence Projections for Scotland (2001-2020)
An Aid to Planning Cancer Services

APPENDIX B: COMPARISON OF RESULTS WITH THOSE PREVIOUSLY PUBLISHED IN CANCER SCENARIOS

The estimates from Cancer Scenarios (old) are compared to those presented here (new) in Table A2.

Table A2: Old and new estimates of number of cases expected in 2010-14 (5 years combined)

Cancer

Old estimates

New estimates

Difference (N)

Difference (%)

Head and Neck

6,738

6,246

-492

-8

Oesophagus

6,223

5,630

-593

-11

Stomach

2,903

3,931

1,028

26

Colorectal

23,254

22,336

-918

-4

Lung

22,577

21,122

-1,455

-7

Pancreas

3,004

3,805

802

21

Melanoma of skin

5,552

5,027

-525

-10

Breast

23,377

21,902

-1,474

-7

Cervix

1,958

1,169

-789

-68

Corpus Uteri

2,051

2,710

658

24

Ovary

3,317

4,013

696

17

Prostate

21,516

13,581

-7,935

-58

Testis

1,361

1,225

-136

-11

Kidney

4,167

4,112

-55

-1

Bladder

9,075

9,196

121

1

Brain

1,974

2,065

91

4

Hodgkin's disease

671

671

0

0

NHL

7,928

6,249

-1,679

-27

Leukaemia

3,752

3,956

204

5

Other and unspecified

17,487

18,239

751

4

Total

168,885

157,185

-11,699

-7

There are two aspects that can account for differences in these estimates compared to those in Cancer Scenarios. First, the population predictions have changed and second, the addition of 5 years of observed data from the cancer registry may impact on the predicted incidence trends.

1) Population predictions

The population projections provided by the Government Actuary Department have changed substantially between those based on Scottish population estimates for 1996 (used for Cancer Scenarios) and their most recent projections, which include information from the 2001 census and based on the Scottish population in 2002 ( see Table A3). The change that will have the largest impact on the cancer predictions is the fact that many more elderly people (aged 75+) are now expected in the Scottish population by 2010-14 (8% more females and 14% more males compared to the 1996-based predictions).

Table A3: Changes in the population projections: 2002-based projections compared to 1996-based projections (which were used in the Cancer Scenarios report)

Sex

Age

Difference (ratio) b2002/b1996

Difference (N) b2002-b1996

2000-04

2005-09

2010-14

2000-04

2005-09

2010-14

F

<35

0.99

0.98

0.97

-79,644

-124,717

-122,823

35-59

1.01

1.01

1.00

34,269

31,876

17,289

60-74

1.02

1.03

1.03

39,515

52,066

47,676

75+

1.03

1.05

1.08

32,269

57,658

74,504

M

<35

0.96

0.95

0.95

-247,260

-247,536

-203,913

35-59

0.98

0.96

0.93

-81,817

-197,813

-238,928

60-74

1.02

1.03

1.03

37,058

54,218

47,272

75+

1.04

1.08

1.14

22,595

53,690

77,519

Source of population projections: Government Actuary Department

2) Incidence trends

The original estimates were based on incidence trends up to 1995, whereas the new estimates include trends up to 2000; a further 5 years of observed data. The underlying trends have changed considerably for some cancers in the last 5 years (e.g. prostate cancer - this is known to be due to changes in diagnostic practice).

REFERENCES

Clayton D, Schliffers E (1987). Models for temporal variations in cancer rates II. Age-period-cohort models. Stat Med; 6: 469-481.

Dyba T (2000). Precision of cancer incidence predictions based on Poisson Distributed Observations. PhD thesis. Finland: University of Helsinki.

Hastie T, Tibshirani R (1990). Generalized Additive Models. London: Chapman and Hall.

New Zealand Ministry of Health (2002). Cancer Projections for New Zealand. Wellington: Ministry of Health.

Osmond C (1985). Using age, period and cohort models to estimate future mortality rates. Int J Epidemiol; 14: 124-9.

Scottish Executive Health Department (2001). Cancer Scenarios: and aid to planning cancer services in Scotland. Edinburgh, Scottish Executive.

Wahba G (1990). Spline models for observational data. Philadelphia. SIAM.

« Previous | Contents |

Page updated: Thursday, June 9, 2005