Screening for prostate cancer in New Zealand general practice

Abstract

Objective

To ascertain the rates and patterns of prostate-specific antigen (PSA) screening in New Zealand men.

Methods

The study population included 35,958 men aged 40+ years, with no prior diagnosis of prostate cancer, enrolled in 31 general practices in the Midland Cancer Network Region of New Zealand in 2010. Computerized practice records were searched for information, including reasons for testing, for men with elevated PSA test results in 2010. PSA results for 2007–2010 were obtained from community laboratories. Screened men were identified and screening rates calculated by age.

Results

Of 9344 men who underwent one or more PSA tests in 2010, 84.9% were classified as having been screened. The overall screening rate was 22.1%, with 24.4% of men aged 70+ years screened. Elevated PSA levels were found in 2.1% of screened men. Of the men screened in 2010, 57.3% had had a screening test in the previous three years.

Conclusions

General practitioners in New Zealand commonly screen men (including those aged 70+) for prostate cancer, despite the lack of trial evidence that these men would benefit from screening. The value of annual PSA testing in men with previous normal PSA levels is unproven. Longer intervals between tests would be appropriate.

INTRODUCTION

In New Zealand (NZ) men, prostate cancer is the most commonly registered cancer and the third most common cause of male cancer deaths. ¹ Prostate-specific antigen (PSA) testing is predominantly carried out by general practitioners. ^2,3 In 2010 there were approximately 350,000 PSA tests carried out for NZ population of 945,000 men aged 40+ years, ⁴ although there is little information on how screening is managed by general practitioners. We used computerized records from general practices, linked with laboratory data, to identify the patterns of PSA screening in a population-based sample.

METHODS

Computerized records from 31 general practices in the Midland Cancer Network Region were used to document the patterns of PSA testing in NZ primary care. Men aged 40+ years who had undergone a PSA test in 2010 were identified by cross-referencing the data from the computerized practice records with laboratory data. This data linkage was based on the National Health Index number, which is a unique identifier for health service users in New Zealand.

Each practice provided baseline data (National Health Index number, date of birth and ethnicity) for all registered male patients aged 40+ years. PSA results for the period from 1 January 2007 to 31 December 2010 were obtained from community laboratories. Data on prostate cancer registrations from 1994 to 2011 were available from the NZ Cancer Registry. We identified 1006 men (2.7%) with a prostate cancer diagnosis prior to 1 January 2010 or prior to any PSA test in 2010 who were excluded from further analysis.

The analysis was based on the overall number of patients, not on the number of tests. In patients with more than one PSA test during 2010, the earliest result was considered in the analysis. When one of the test results (not necessarily the earliest one) was elevated, the man was categorized as having an elevated PSA test during 2010. PSA values were classified as being elevated when they exceeded the age-specific levels recommended by local laboratories (adapted from Oesterling et al. ⁵ ).

The practice records of men with an elevated PSA test in 2010 were searched for information on the reasons for testing. Men were classified as having been screened when the reason for PSA testing was screening (in men with elevated PSA test in 2010), and when there were no elevated PSA tests recorded in the three years prior to 2010 (in men without elevated PSA test in 2010).

The access to computerized records of general practices was obtained through close cooperation with general practitioners and practice managers. Ethical approval was issued by the Northern Y Ethics Committee (NTY/11/02/019).

RESULTS

We identified 35,958 male patients aged 40+ registered with 31 general practices in the Midland Cancer Network Region in 2010 who had had no previous history of prostate cancer.

In total, 9344 men had a PSA test in 2010, of whom 84.9% were considered to be asymptomatic screened men. The remaining 1408 men had a previously elevated PSA test result between 2007 and 2009, a record of previous prostate issues, or symptoms at the time of the first PSA test in 2010.

Table 1 shows the age-specific screening rates, proportions of men with elevated PSA and those with normal test results between 2007 and 2009. The overall screening rate was 22.1% (7936/35,958). The highest screening rates were observed in men aged 60–69 (31.5%), followed by men aged 70–79 (27.7%). Only 2.1% of men had an elevated PSA test in 2010. Fifty-seven percent of men screened in 2010 had at least one normal PSA result in the preceding three years. Men without a PSA test result between 2007 and 2009 were more likely to have an elevated PSA test compared with those previously tested (3.1% versus 1.4%; Fisher's exact test P < 0.0001).

Table 1

Proportions of men with screening PSA test in 2010, elevated PSA level and with a PSA test between 2007 and 2009 by age group

		Men screened in 2010		Men with elevated PSA (from screened)		Men with PSA test (2007–2009) (from screened)
Age group (years)	N	n	%	n	%	n	%
40–49	11752	1434	12.2	21	1.5	424	29.6
50–59	10466	2604	24.9	43	1.7	1385	53.2
60–69	7638	2407	31.5	71	2.9	1667	69.3
70–79	4312	1193	27.7	27	2.3	888	74.4
80+	1790	298	16.6	8	2.7	183	61.4
Total	35958	7936	22.1	170	2.1	4547	57.3

DISCUSSION

The Urological Society of Australia and New Zealand ⁶ recommends that asymptomatic men aged 55–69 years should be offered a PSA test and digital rectal examination after they have been advised on the risks and benefits of PSA screening. In contrast, the National Screening Advisory Committee of the NZ Ministry of Health does not support population-based screening. ⁷ The US Preventive Task Force has recently advised against PSA screening regardless of patient's age. ⁸

PSA testing has been increasingly used as a screening test for prostate cancer detection, despite its limitations. ⁹ In our study, approximately every fourth man aged over 40 was screened in primary care during a one-year period. This result is consistent with a questionnaire-based NZ study, which reported screening rates of 18% in 2008 and 22% during the years 2003–2007. ³ High screening rates may be a reflection of the privatized nature of NZ general practice, where state support only partially covers health care costs. Results from our pilot study ² indicated that 70% of the screening tests were initiated by the general practitioner.

In our study, 24% of men aged over 70 had been screened. While it may be argued that there is benefit from screening in men aged under 70, in older men there is no evidence of reduced mortality from randomized controlled trials. ¹⁰

Men in our study were frequently tested, with more than half of the men screened in 2010 having one or more PSA tests with a normal result in the preceding three years. We found that elevated PSA test results were significantly more commonly detected in screened men with no previous tests compared with those tested prior to 2010. It would seem prudent, therefore, to recommend less frequent screening in the NZ general practice.

CONCLUSIONS

We examined PSA screening rates in NZ men by linking computerized records from general practices to laboratory data. We found that New Zealand general practitioners commonly screen men, including men aged over 70, for whom there is no trial evidence of benefit from screening. We recommend that general practitioners should be discouraged from initiating screening in asymptomatic men over the age of 70. We also question the value of annual PSA testing in men with previous normal PSA levels, and suggest that longer intervals between tests would be appropriate.