Oncological outcomes in an Australian cohort according to the new prostate cancer grading groupings.

BACKGROUND: A new 5-tiered grading grouping system has recently been endorsed for reporting of prostate cancer (PCa) grade to better reflect escalating risk of progression and cancer death. While several validations of the new grade groupings have been undertaken, most have involved centralised pathological review by specialist urological pathologists. METHODS: Participants included 4268 men with non-metastatic PCa diagnosed between 2006 and 2013 from the multi-institutional South Australia Prostate Cancer Clinical Outcomes Collaborative registry. PCa-specific survival and biochemical recurrence-free survival were compared across the five grade groups using multivariable competing risk regression. RESULTS: For the entire cohort, risk of PCa death increased with increasing grade groups (at biopsy) Adjusted subdistribution-hazard ratios [sHR] and 95% confidence intervals [95%CI] were: 2.2 (1.5-3.6); 2.5 (1.6-4.2); 4.1 (2.6-6.7) and 8.7 (4.5-14.0) for grade groups II (pattern 3 + 4), III (pattern 4 + 3), IV (total score 8) and V (total score 9-10) respectively, relative to grade group I (total score < =6). Clear gradients in risk of PCa death were observed for radical prostatectomy (RP), but were less clear for those who had radiotherapy (RT) with curative intent and those who were managed conservatively. Likewise, risk of biochemical recurrence increased across grade groups, with a strong and clear gradient for men undergoing RP [sHR (95%CI): 2.0 (1.4-2.8); 3.8 (2.9-5.9); 5.3 (3.5-8.0); 11.2 (6.5-19.2) for grade groups II, III, IV and V respectively, relative to grade group I], and a less clear gradient for men undergoing RT. CONCLUSION: In general, the new five-tiered grade groupings distinguished PCa survival and recurrence outcomes for men with PCa. The absence of a clear gradient for RT may be due to heterogeneity in this patient group.

Realising opportunities for evidence-based cancer service delivery and research: linking cancer registry and administrative data in Australia.

The traditional roles of Australian cancer registries have been incidence, mortality and survival surveillance although increasingly, roles are being broadened to include data support for health-service management and evaluation. In some Australian jurisdictions, cancer stage and other prognostic data are being included in registry databases and this is being facilitated by an increase in structured pathology reporting by pathology and haematology laboratories. Data linkage facilities are being extended across the country at national and jurisdictional level, facilitating data linkage between registry data and data extracts from administrative databases that include treatment, screening and vaccination data, and self-reported data from large population cohorts. Well-established linkage protocols exist to protect privacy. The aim is to gain better data on patterns of care, service outcomes and related performance indicators for health-service management and population health and health-services research, at a time of increasing cost pressures. Barriers include wariness among some data custodians towards releasing data and the need for clearance for data release from large numbers of research ethics committees. Progress is being made though, and proof of concept is being established.

A systematic review of surgical biopsy for LCIS found at core needle biopsy - do we have the answer yet?

BACKGROUND: The natural history of lobular carcinoma in-situ (LCIS) suggests that women are at increased risk of subsequent invasive breast cancer. Questions of effective management for women with this lesion have led to the need for evidence-based guidance and, in particular, guidance regarding management after LCIS is found at core needle biopsy (CNB). METHODS: A systematic review was conducted to determine the most appropriate management for women with LCIS found at CNB. A comprehensive search of the scientific literature was conducted to identify the literature pertaining to this population. Critical appraisal of the literature, data extraction and a narrative synthesis of the results were conducted. The outcome of interest was upgrade of diagnosis to invasive breast cancer or ductal carcinoma in-situ (DCIS). RESULTS: Sparse data, with limited generalisability and considerable uncertainty, are available for women with LCIS at CNB. Nine studies were identified that met pre-specified inclusion criteria. The reported estimates of upgrade of diagnosis from LCIS to invasive breast cancer or DCIS ranged from 2% to 25%. The body of evidence was limited by its retrospective nature, risk of selection bias and poor generalisability to all women with LCIS at CNB. Further, higher quality research is required to determine the best approach for women with LCIS at CNB with any certainty.

What factors affect the odds of NSW cancer patients presenting with localised as opposed to more advanced cancer?

OBJECTIVE: The purpose was to examine the odds of presenting with localised as opposed to more advanced cancer by place of residence to gain evidence for planning early detection initiatives. METHODS: Design, settings and participant's cases of invasive cancer reported to the NSW population-based Cancer Registry for the 1980-2008 diagnostic periods. Main outcome measure(s) between 1980 and 2008, 293,848 of reported cases (40.2%), had localised cancer at diagnosis. Logistic regression analysis was undertaken to determine the odds of localised cancer by place of residence for all cancers sites combined while adjusting for age, sex, period of diagnosis, socioeconomic status, migrant status and prognosis (as inferred from cancer type). RESULTS: Multivariate logistic regression analysis indicated that patients from rural areas were less likely than urban patients to present with localised cancer after adjusting for other socio-demographic factors and prognosis by cancer type (regardless of how rurality was classified). The difference ranged from 4% for remote (OR = 0.96, 95% CI 0.95-0.98) to 14% (OR = 0.86, 95% CI 0.79-0.84) for very remote compared with highly accessible areas. It is estimated that a maximum of 4,205 fewer cases of localised cancer occurred in patients from rural areas over the study period than expected from the stage distribution for urban patients. Residents aged between 30 and 74 years of age at diagnosis and those living in high socioeconomic status areas were more likely to present with localised cancer. By contrast, people aged 75 years or older at diagnosis, migrants from non-English-speaking countries and people diagnosed in more recent diagnostic periods were less likely to present with localised cancer. CONCLUSIONS: Targeted strategies that specifically encourage earlier diagnosis and treatment that may subsequently influence better survival are required to increase the proportion of NSW residents presenting with localised cancer at diagnosis.

Clinical and pathological factors predictive of lymph node status in women with screen-detected breast cancer.

Two thousand one hundred and thirty five asymptomatic invasive breast cancers detected through screening mammography were analysed to identify predictors of lymph node involvement. Multivariable analysis indicated that predictors included larger tumour diameter, an infiltrating ductal or lobular histological type, multifocal disease, a palpable lesion, and a younger age at diagnosis. An association also was found between nodal involvement and the presence of an extensive in situ component (EIC). Grade was associated with nodal involvement as a univariate predictor. It would be more accurate for screening assessment clinics to use models for predicting nodal status that were customised to their own experience rather than generic models developed in other settings that related predominantly to symptomatic cancer. These models could assist clinical decision-making on axillary node dissection and give guidance to pathologists on numbers of tissue sections to examine.

Treatment and survival from colorectal cancer: the experience of patients at South Australian teaching hospitals between 1980 and 2002.

AIMS: To evaluate trends in colorectal cancer survival and treatment at South Australian teaching hospitals and degree of adherence to treatment guidelines which recommend adjuvant chemotherapy for Dukes' C colon cancers and combined chemotherapy and radiotherapy for high-risk rectal cancers. MATERIALS AND METHODS: Trends in disease specific survival and primary treatment were analysed, and comparisons drawn between diagnostic epochs, using cancer registry data from South Australian teaching hospitals. Statistical methods included univariate and multivariable disease specific survival analyses. RESULTS: Five-year survival increased from 48% in 1980-1986 to 56% in 1995-2002. Largest gains were for stage C, where survivals were higher when chemotherapy was part of the primary treatment. By comparison, gains in 1-year survival were largest for stage D. Chemotherapy was provided for 4% of patients with colorectal cancers in 1980-1986, increasing to 32% in 1995-2002. Among stage C cases below 70 years at diagnosis, the proportion having chemotherapy increased to 83% in 1995-2002. The most common chemotherapy was fluorouracil (5FU) as a single agent in 1980-1986 and 5FU with leucovorin in 1995-2002. As expected, radiotherapy was used more frequently for rectal than colon cancers, and particularly for stage C. Among stage C rectal cases below 70 years, the proportion having radiotherapy increased from 10% in 1980-1986 to 57% in 1995-2002. Approximately 93% of colorectal cancers were treated surgically. Patients not treated surgically tended to be aged 80 years or more and to present with distant metastases. CONCLUSIONS: Trends in chemotherapy and radiotherapy accord with evidence-based recommendations. There have been reassuring gains in survivals after adjusting for stage, grade and other prognostic indicators. The data show survival gains and treatment patterns that individual hospitals can use as benchmarks when evaluating their own experience.

Detection by screening mammography is a powerful independent predictor of survival in women diagnosed with breast cancer.

Four hundred and sixteen invasive breast cancers, detected initially by mammography, were compared with 929 presenting symptomatically, all treated at a South Australian teaching hospital. Predictable differences included lower stages and grades, less vascular invasion and proliferative activity, and more hormone-receptor expression among the mammographically detected. Unpredicted differences included significantly higher survivals for mammographically detected cases throughout the 9 year follow-up period after adjusting for stage and the Nottingham Prognostic Index. In a multivariable analysis, differences in stage, grade, and hormone receptor expression accounted for only about half the survival advantage of mammographically detected tumours. Accounting for additional person and tumour characteristics had only a marginal effect on this result. This suggests that detection by mammography has independent favourable prognostic significance beyond that explained by conventional indicators. If confirmed, this finding would have important implications for the prognostic advice given to women and may merit further investigation into its underlying biological mechanisms.

The coverage of cancer patients by designated palliative services: a population-based study, South Australia, 1999.

Our aims were to determine the extent of coverage by designated palliative care services of the population of terminally ill cancer patients in South Australia, and to identify the types of patients who receive these services and the types who do not. All designated hospice and palliative care services in South Australia notified to the State Cancer Registry the identifying details of all their patients who died in 1999. This information was cross-referenced with the data for all cancer deaths (n=3086) recorded on the registry for 1999. We found that the level of coverage by designated palliative services of patients who died with cancer in 1999 was 68.2%. This methodology was previously used to show that the level of coverage had increased from 55.8% for cancer deaths in 1990 to 63.1% for those in 1993. Patients who died at home had the largest coverage by palliative services (74.7%), whereas patients who died in nursing homes had the lowest coverage (48.4%). Patients who did not receive care from these palliative services tended to be 80 years of age or older at death, country residents, those with a survival time from diagnosis of three months or less, and those diagnosed with a prostate, breast, or haematological malignancy. Gender, socioeconomic status of residential area, and race were not related to coverage by a designated palliative service, whereas migrants to Australia from the UK, Ireland, and Southern Europe were relatively high users of these services. We conclude that the high level of palliative care coverage observed in this study reflects widespread support for the establishment of designated services. When planning future care, special consideration should be given to the types of patients who most miss out on these services.

An analysis of factors associated with interval as opposed to screen-detected breast cancers, including hormone therapy and mammographic density.

A comparison of 270 interval breast cancers in South Australian women aged 50 years and over with 628 age-matched screen-detected cases indicated that the former had more advanced stages (P<0.001), higher grades (P<0.001), and more frequently a history of past breast problems (P<0.027). After adjusting for these factors, and presence of a self-reported breast lump at the last screen, using conditional logistic regression, hormone replacement therapy (HRT) exposure in the 6 months prior to this screen had a raised relative odds (95% CL) of an interval cancer of 1.48 (1.02, 2.14). For 479 women where breast density was measured, high density showed an elevated relative odds of an interval cancer of 2.62 (1.71, 4.02). The relative odds of a high density was raised to 2.02 (1.33, 3.06) when the HRT history was positive. Screeners should be aware when there is a history of HRT or past breast problems, or a high breast density, that there is an increased probability of a subsequent interval cancer.

Tumour location and prognostic characteristics as determinants of survival of women with invasive breast cancer: South Australian hospital-based cancer registries, 1987-1998.

Survivals from breast cancer varied by location of lesion (P<0.001), with 10-year survivals of 61% applying for central (n=772), 73% for medial (n=350), and 72% for lateral (n=966) lesions. Univariate analyses of determinants of central locations indicated that the following were predictive: a more advanced TNM stage (P<0.001); a larger tumour diameter (P=0.002); a higher grade (P=0.032); a negative oestrogen receptor status (P=0.004); a negative progesterone receptor status (P=0.004); and histological type (P=0.011), with more of the lobular lesions being located centrally. Cox proportional hazards regression indicated that the relative risk (95% confidence limits) of case fatality for central, as opposed to other, lesions reduced from 1.46 (1.20, 1.78) to 1.16 (0.95, 1.41) when stage was added to the model, with no other factor having an additional conditioning effect. It is concluded that central lesions have worse outcomes, mostly due to their more advanced stages. Means of finding these tumours earlier should be investigated.

Use of fake tanning lotions in the South Australian population.

OBJECTIVE: To explore the relationship between the use of fake tanning lotions and repeated sunburn among South Australian adults, with a view to informing the Anti-Cancer Foundation of South Australia's (ACFSA) policy on fake tanning products. STUDY DESIGN: Population survey. PARTICIPANTS: 2005 South Australians aged 18 years or older, selected randomly from the electronic White Pages. MAIN OUTCOME MEASURES: Self-reported use of fake tanning lotions in the past 12 months; frequency of sunburn over summer; and various sun-protective behaviours. RESULTS: 2,005 of the 2,536 eligible participants (79%) were surveyed by telephone. Fake tan use was most prevalent among women (15.9%), people aged 18-24 years (15.4%), and people with household incomes above $40 000 per year (11.9%). Fake tan users were more likely than non-users to use sunscreens (81.3% v 56.5%; P < 0.001), but less likely to take other precautions such as wearing hats (40.9% v 51.0%; P = 0.04) and protective clothing (22.3% v 34.1%; P = 0.005). They were also more likely to report having been burnt more than once over summer (26.2% v 16.5%; P = 0.025). Multivariate analysis indicates a statistically significant association between fake tan use and repeated sunburn (odds ratio, 2.07; 95% confidence interval, 1.17-3.69), which was independent of age, sex, skin type and sun-protection practices. CONCLUSION: Users of fake tanning products may be at greater risk of repeated sunburn. The ACFSA sees no justification at this stage for altering its present policy position of not actively promoting the use of fake tanning lotions as a means of reducing sunburn.

Where patients with cancer die in South Australia, 1990-1999: a population-based review.

OBJECTIVE: To determine the place of death of South Australians who die of cancer. DESIGN: A population-based, cross-sectional study of data from the South Australian Cancer Registry. PARTICIPANTS: 29,230 patients with cancer dying in 1990-1999. MAIN OUTCOME MEASURES: Place of death; patient demography; year of death; survival from diagnosis; and type of cancer. RESULTS: 25.0% of patients died in a metropolitan public hospital, 19.9% in a hospice, 16.9% in a country hospital, 15.8% at a private residence, 12.7% in a metropolitan private hospital, and 9.7% in a nursing home. Although the change in place of death was not marked, multivariate logistic regression showed a secular trend away from metropolitan public hospitals towards metropolitan private hospitals and, in 1998-1999, towards nursing homes. Patients dying of cancer in a metropolitan public hospital were more likely to be younger, males, born outside Australia, and residents of lower socioeconomic areas of Adelaide. They were also more likely to have died within three months of diagnosis, and to have a haematological malignancy or a cancer of the upper digestive tract, lung or female breast. In contrast, patients dying at a private residence tended to be under 70 years and comprise longer-term survivors. Country residents were less likely than Adelaide residents to die in a hospice. CONCLUSION: The proportion of patients dying in different settings have health service implications. The relatively low use of hospice facilities by country patients may reflect differences in access to hospice facilities.

Trends in hospital readmission for asthma: has the Australian National Asthma Campaign had an effect?

OBJECTIVES: To describe patterns of hospital readmission for asthma in South Australia from 1989 to 1996, in relation to implementation of the National Asthma Campaign. DESIGN AND SETTING: A comparison of hospital admissions in South Australia of patients aged between one year and 49 years for three conditions: asthma (or respiratory failure with asthma as an underlying condition) and two control conditions--diabetes and epilepsy. Individuals were identified by Medicare number and date of birth. OUTCOME MEASURES: Hospital readmission within 28 days and within one year. RESULTS: Overall, by 1996, there was a statistically significant decline in the risk of readmission for asthma within 28 days of 18% and within one year of 17% compared with 1989 readmission rates. There were no reductions in the risk of readmission for diabetes or epilepsy, suggesting that the decline in risk of readmission for asthma was greater than the underlying effects of general changes in hospital casemix. CONCLUSIONS: The decline in risk of readmission may reflect changes in asthma severity or improved management practices. However, hospital readmission rates still remain high, and to further reduce readmissions for asthma there is a need to identify factors related to presentation for asthma at accident and emergency departments.

Rehospitalisation for asthma within 12 months: unequal rates on the basis of gender in two hospitals.

BACKGROUND: High asthma morbidity has been reported in certain demographic groups in geographical areas of low socioeconomic status (SES). AIM AND METHODS: We tested for an ecological association between the gender of people being frequently hospitalised for asthma and the SES of the area in which they lived, using a cross-sectional study. RESULTS: Women represented 75% of the readmission population at The Queen Elizabeth Hospital (TQEH--low SES hospital) and 55% at Modbury Hospital (moderate-high SES hospital). Women at TQEH were significantly more likely to have one readmission within 12 months and over 30 times more likely to have two or more readmissions than women at Modbury Hospital. CONCLUSION: The ecological association observed in this study needs to be confirmed elsewhere in Australia.

Near-fatal asthma in South Australia: descriptive features and medication use.

BACKGROUND: Self-reported prior morbidity levels and medication use among survivors of a near-fatal asthma attack (NFA) were studied. AIMS: To identify deficiencies in asthma management and opportunities for intervention. METHODS: A hundred and twenty-seven consecutive patients aged 15 years or more presenting with a NFA to accident and emergency departments of teaching hospitals were interviewed. RESULTS: High levels of morbidity due to asthma were reported. Most cases (79%) reported symptoms occurring at least weekly in the three months before their NFA. A mean of 20.8 days was reportedly lost from work, school or other usual daily activity in the 12 months before these events. Regular use of beta agonist as nebuliser solution was reported by 27% of cases, increasing to 34.5% in response to increased symptoms, while 41% reported use of nebulised beta agonist in response to the NFA event. Less than half of all cases (46%) reported using an inhaled corticosteroid on a regular basis. Oral corticosteroids were used by 33% of cases at times of increased symptoms in the preceding 12 months. However, only 7% of cases reported initiating or increasing oral corticosteroids at the time of the NFA. CONCLUSIONS: Despite high levels of prior asthma morbidity, regular preventive inhaled corticosteroid use was not widespread in this series of NFA asthmatics. By comparison, over-reliance on regular beta agonist medication was common. Oral corticosteroids were rarely commenced in response to the NFA.

Medroxyprogesterone acetate therapy in advanced breast cancer: the predictive value of androgen receptor expression.

PURPOSE: To determine the predictive value of androgen receptor (AR) levels in primary tumors of women who undergo medroxyprogesterone acetate (MPA) therapy for advanced breast cancer after relapse on tamoxifen adjuvant therapy. METHODS: Between 1984 and 1987 at Flinders Medical Centre, South Australia, 136 postmenopausal women received adjuvant tamoxifen therapy for lymph node-positive breast cancer. Estrogen receptor (ER), progesterone receptor (PgR), and AR levels, tumor size, and degree of axillary node involvement were determined at the time of diagnosis. The median follow-up period was 81 months; 89 women developed metastatic disease, 83 of whom subsequently received MPA (500 mg/d). The objective response rate ([RR] ie, complete response [CR] and partial response [PR]) and progression-free interval (PFI) were assessed in response to MPA therapy. Associations between RR, PFI, and primary tumor characteristics including ER, PgR, and AR levels were examined using the Mann-Whitney U test, Kaplan-Meier product-limit estimator, and Cox proportional hazards regression, as appropriate. RESULTS: Thirty-two of 83 patients (38.6%) responded to MPA. RR was significantly associated with the presence of AR (P < .001), but not with other primary tumor characteristics or duration of tamoxifen therapy. After initiation of MPA treatment, PFI increased with increasing concentration of AR in the primary tumor. CONCLUSION: Response to MPA after adjuvant tamoxifen treatment for lymph node-positive breast cancer was positively associated with AR level in the primary tumor. This finding suggests that MPA action in breast cancer may be mediated in part by the AR.

Characteristics of near-fatal asthma in childhood.

As part of the South Australian asthma mortality survey, we examined 30 cases of near-fatal asthma attacks in children under 15 years of age who were seen over a 3-year period from May 1988 to June 1991. Subjects presented with asthma and either respiratory arrest, PaCO2 above 50 mm Hg, and/or an altered state of consciousness or inability to speak on presentation at a metropolitan Adelaide teaching hospital. A standardized interview and questionnaire was completed with subjects/parents and medical practitioners. Data were reviewed by the assessment panel which made collective judgments based on predetermined criteria. Seventeen patients (57%) were male, 20% were less than 7 years of age, and the majority (53%) were aged between 12 and 15 years. The majority (83%) had severe asthma and only one case (3.3%) had mild asthma. Half of the subjects were waking every night due to asthma and 79% had significant exercise limitation. A quarter of the subjects had a previous ICU admission and 70% had a hospital admission in the last 12 months. Primary care was carried out by a general practitioner in 57% of cases, and 70% of subjects had a crisis plan. Only 46% of those older than 7 years of age had ever used a peak-flow meter. Eighty percent of subjects or their families had high denial scores, and in 73% of cases psychosocial factors were considered to be significant. Eighty percent of cases experienced acute progressive respiratory distress, and 63% of cases delayed seeking medical care.(ABSTRACT TRUNCATED AT 250 WORDS)

Trends in prostate cancer incidence and mortality rates in South Australia, 1977-1993.

OBJECTIVE: To investigate trends in recorded incidence and mortality rates for prostate cancer in South Australia. DESIGN: A multiple Poisson regression analysis of recorded incidence (by diagnostic period) and mortality (by year of death), after adjusting for age at diagnosis and residential location. SUBJECTS AND SETTING: 8073 patients with prostate cancer and 2659 who died of prostate cancer as notified to the South Australian Cancer Registry for 1977-1993. MAIN OUTCOME MEASURES: The relative risk of a recorded diagnosis of prostate cancer (by period of diagnosis), and of a death from prostate cancer (by year of death). RESULTS: During 1977-1989, the recorded age-standardised incidence of prostate cancer was stable, but it increased markedly thereafter. The relative risk (95% confidence limits) of diagnosed prostate cancer was 1.36 (1.29, 1.43) in 1990-1992, and 2.26 (2.12, 2.42) in 1993, when compared with 1977-1989. There was a smaller and less certain increase in prostate cancer mortality. CONCLUSIONS: The large increase in recorded incidence of prostate cancer in South Australia is thought to be due mostly to increased disclosure of latent cases from increased clinical investigations. Until there is experimental evidence of health benefits from screening and related investigations for prostate cancer in asymptomatic men, it will be difficult to reconcile benefits with costs.

Murray River water, raised cyanobacterial cell counts, and gastrointestinal and dermatological symptoms.

OBJECTIVE: To investigate whether exposure to Murray River and allied water sources during a period of raised cyanobacterial cell counts was associated with gastrointestinal and dermatological symptoms. DESIGN: A case-control study selecting gastrointestinal and dermatological cases and controls from subjects attending 21 general practitioners in eight Murray River towns. The association between the proportion of consultations for such symptoms and mean log cyanobacterial count was also examined. SUBJECTS: 102 gastrointestinal cases, 86 dermatological cases and 132 controls. MAIN OUTCOME MEASURE: The relative odds of gastrointestinal and dermatological symptoms, respectively, as opposed to no such symptoms, according to water-contact history during the week preceding the medical consultation. RESULTS: After adjusting for concurrent risk factors, subjects drinking chlorinated river water rather than rain water had a raised risk of gastrointestinal symptoms (P = 0.008), and those using untreated river water for domestic purposes rather than rain water had a raised risk of gastrointestinal (P = 0.034) and of dermatological (P = 0.048) symptoms. The proportion of consultations for gastrointestinal and dermatological symptoms correlated on a weekly basis with the mean log cyanobacterial cell count, although statistical significance was not achieved for the correlation with dermatological consultations or for separate reaches of the river. CONCLUSIONS: The raised risks of gastrointestinal and dermatological symptoms in those using Murray River water for drinking and other domestic purposes are consistent with causal relationships. However, the evidence for adverse health effects is, at best, only suggestive. Further research is indicated.

Trends in prognostic factors of melanoma in South Australia, 1981-1992: implications for health promotion.

OBJECTIVE: To analyse trends in prognostic factors at diagnosis of melanoma to indicate targets for health promotion. METHODS: Cases of melanoma notified to the South Australian Cancer Registry during 1981-1992 (1361 in situ and 4509 invasive) were analysed by tumour site, histological type, stage, thickness, period of diagnosis and sociodemographic characteristics of the patient. RESULTS: The relative probability of being diagnosed with an in-situ, rather than a thin invasive, lesion increased progressively between 1984-1986 and 1990-1992. The relative probability of having an in-situ melanoma diagnosed was: lower among those aged 80 years or more; higher among patients residing in the upper socioeconomic areas of Adelaide; high for lesions on the face; and high for lentigo maligna as opposed to superficial spreading lesions. The relative probability of invasive lesions being diagnosed when more than 1.5 mm thick decreased progressively between the 1981-1983 and 1990-1992 diagnosis periods. The relative probability of thick lesions being diagnosed was higher among older patients and for nodular lesions compared with superficial spreading or lentigo maligna lesions. Facial lesions and those in males tended to be thicker. CONCLUSIONS: There are favourable trends towards earlier detection of melanomas in South Australia, probably reflecting the effect of skin-cancer awareness campaigns. Further campaigns should emphasise early detection in people aged 60 years and over, males and residents of less affluent areas. Early detection of nodular melanomas and invasive lesions of the face also warrants special attention.