Serum inflammatory markers in relation to prostate cancer severity and death in the Swedish AMORIS study.

Inflammation is a well-documented driver of cancer development and progression. However, little is known about its role in prostate carcinogenesis. Thus, we examined the association of C-reactive protein (CRP), haptoglobin, albumin and white blood cells (WBC) with prostate cancer (PCa) severity (defined by PCa risk category and clinicopathological characteristics) and progression (defined by PCa death). We selected 8,471 Swedish men with newly diagnosed PCa who had exposure measurements taken approximately 14 years prior to diagnosis. We calculated odds ratio (OR) and 95% confidence interval (CI) for the associations between the inflammatory markers and PCa severity using logistic regression, while Cox proportional hazard regression was used for the associations with overall and PCa death. Serum CRP levels were associated with increased odds of high risk and metastatic PCa, and high PSA levels (≥20 µg/L) (OR: 1.29; 95% CI: 1.06-1.56, 1.32; 1.05-1.65 and 1.51; 1.26-1.81, respectively). Similarly, higher haptoglobin levels were associated with increased odds of metastatic PCa, high PSA level and possibly high grade PCa (1.38; 1.10-1.74, 1.50; 1.17-1.93 and 1.25; 1.00-1.56, respectively). Albumin was positively associated with Gleason 4 + 3 tumour (1.38; 1.02-1.86) and overall death (HRunit increase in log : 1.60; 95% CI: 1.11-2.30), but inversely associated with high risk PCa and high PSA levels (≥20 µg/L) (0.71; 0.56-0.89 and 0.72; 0.5 9-0.90). WBC was associated with increased odds of T3-T4 PCa. Except for albumin, none of these markers were associated with PCa death or overall death. Systemic inflammation as early as 14 years prior to diagnosis may influence prostate cancer severity.

Determinants of non-adherence to adjuvant endocrine treatment in women with breast cancer: the role of comorbidity.

PURPOSE: To examine factors associated with non-adherence during 5 years of endocrine treatment, including the possible influence of comorbidity burden and specific medical conditions. METHODS: From all women diagnosed with stage I-III, ER-positive breast cancer in Stockholm-Gotland, Uppsala-Örebro and Northern Sweden between 2006 and 2009, we included 4645 women who had at least one dispensation of tamoxifen or aromatase inhibitors (AIs) and 5 years of follow-up without distant recurrence. A medical possession ratio of < 80% was used to define non-adherence. Logistic regression was used to estimate odds ratios (ORs) and 95% confidence intervals (CIs) of non-adherence. RESULTS: During follow-up, 977 (21%) women became non-adherents. Non-adherence was associated with greater comorbidity burden assessed by Charlson comorbidity index (CCI) during follow-up (OR 1.43; 95% CI 1.08-1.88 for ≥ 2 additional scores compared to 0), pre-diagnostic HRT use (OR 1.99; 1.58-2.49), not married (OR 1.42; 1.23-1.64), high educational level (OR 1.25; 1.02-1.53 compared to lowest level), and use of symptom-relieving drugs. HER-2 positivity (OR 0.61; 0.45-0.81) and adjuvant chemotherapy (OR 0.42; 0.35-0.52) were associated with lower odds of non-adherence. Similar patterns were observed for the presence of lymph node metastasis, higher tumour grade, and use of AIs compared to tamoxifen. Myocardial infarction and chronic pulmonary disease was suggested as leading conditions associated with non-adherence in women with increasing CCI. CONCLUSION: We identified subgroups of women with breast cancer at increased risk of non-adherence. Our findings related to comorbidity suggest the importance of focusing on the presence of specific co-existing conditions when monitoring adherence.

Review of 103 Swedish Healthcare Quality Registries.

BACKGROUND AND OBJECTIVES: In the past two decades, an increasing number of nationwide, Swedish Healthcare Quality Registries (QRs) focusing on specific disorders have been initiated, mostly by physicians. Here, we describe the purpose, organization, variables, coverage and completeness of 103 Swedish QRs. METHODS: From March to September 2013, we examined the 2012 applications of 103 QRs to the Swedish Association of Local Authorities and Regions (SALAR) and also studied the annual reports from the same QRs. After initial data abstraction, the coordinator of each QR was contacted at least twice between June and October 2013 and asked to confirm the accuracy of the data retrieved from the applications and reports. RESULTS: About 60% of the QRs covered ≥80% of their target population (completeness). Data recorded in Swedish QRs include aspects of disease management (diagnosis, clinical characteristics, treatment and lead times). In addition, some QRs retrieve data on self-reported quality of life (EQ5D, SF-36 and disease-specific measures), lifestyle (smoking) and general health status (World Health Organization performance status, body mass index and blood pressure). CONCLUSION: Detailed clinical data available in Swedish QRs complement information from government-administered registries and provide an important source not only for assessment and development of quality of care but also for research.

Cancer survival in Australia, Canada, Denmark, Norway, Sweden, and the UK, 1995-2007 (the International Cancer Benchmarking Partnership): an analysis of population-based cancer registry data.

BACKGROUND: Cancer survival is a key measure of the effectiveness of health-care systems. Persistent regional and international differences in survival represent many avoidable deaths. Differences in survival have prompted or guided cancer control strategies. This is the first study in a programme to investigate international survival disparities, with the aim of informing health policy to raise standards and reduce inequalities in survival. METHODS: Data from population-based cancer registries in 12 jurisdictions in six countries were provided for 2·4 million adults diagnosed with primary colorectal, lung, breast (women), or ovarian cancer during 1995-2007, with follow-up to Dec 31, 2007. Data quality control and analyses were done centrally with a common protocol, overseen by external experts. We estimated 1-year and 5-year relative survival, constructing 252 complete life tables to control for background mortality by age, sex, and calendar year. We report age-specific and age-standardised relative survival at 1 and 5 years, and 5-year survival conditional on survival to the first anniversary of diagnosis. We also examined incidence and mortality trends during 1985-2005. FINDINGS: Relative survival improved during 1995-2007 for all four cancers in all jurisdictions. Survival was persistently higher in Australia, Canada, and Sweden, intermediate in Norway, and lower in Denmark, England, Northern Ireland, and Wales, particularly in the first year after diagnosis and for patients aged 65 years and older. International differences narrowed at all ages for breast cancer, from about 9% to 5% at 1 year and from about 14% to 8% at 5 years, but less or not at all for the other cancers. For colorectal cancer, the international range narrowed only for patients aged 65 years and older, by 2-6% at 1 year and by 2-3% at 5 years. INTERPRETATION: Up-to-date survival trends show increases but persistent differences between countries. Trends in cancer incidence and mortality are broadly consistent with these trends in survival. Data quality and changes in classification are not likely explanations. The patterns are consistent with later diagnosis or differences in treatment, particularly in Denmark and the UK, and in patients aged 65 years and older. FUNDING: Department of Health, England; and Cancer Research UK.

Anti-androgen prescribing patterns, patient treatment adherence and influencing factors; results from the nationwide PCBaSe Sweden.

PURPOSE: Adherence has not been studied in male oncology populations. The aim of this study on both the prescriber and user perspectives in prostate cancer treatment was to analyse real-life prescribing patterns of anti-androgens (AA), primarily bicalutamide, and factors influencing the patients' adherence to treatment. METHODS: A nationwide clinical cohort of incident prostate cancer, PCBaSe, was linked to the Swedish Prescribed Drug Register. Men with a planned first line monotherapy AA treatment were identified; dosages and extent of off-label treatment were investigated. Cumulative incidence proportions for reasons for drug discontinuation were calculated. Factors potentially influencing adherence were explored using the medical possession ratio based on the individual prescribed daily dose. RESULTS: First line monotherapy AA was planned in 4.4 % of all incident cases and in 2.1 % of low risk disease cases. Among 1,406 men prescribed bicalutamide, 1,109 (79 %) received the approved daily dose of 150 mg. Discontinuation reasons differed with disease severity. Off-label, low-dose prescription associated with age above 75 years and disease categorised as low risk was noted in 297 men (21 %). Sixty percent of the men adhered well, i.e. to ≥80 %. Age above 75 years and less severe disease were both negatively associated with adherence. CONCLUSIONS: Patient age and tumour risk group influenced the prescriber's choice of dose, pointing to important issues for critical reflection. Possible over-treatment was noted in low risk disease. Interventions to increase adherence in older men and in men with less severe disease are worth considering after critically reviewing the appropriateness of the treatment indication, especially in the latter case.

Excess mortality for men and women above age 70 according to level of care during the first wave of COVID-19 pandemic in Sweden: A population-based study.

BACKGROUND: Both age and comorbidity are established risk factors for death among those infected with COVID-19. Because they often co-exist, it is difficult to assess if age is a risk factor on its own. METHODS: We used administrative register data of the total Swedish population from 01/2015 until 07/2020. We stratified the population aged 70+ into three groups according to level of care (in care homes, with home care, and in independent living). Within these groups, we explored the level of excess mortality in 2020 by estimating expected mortality with Poisson regression and compared it to observed levels. We investigated if excess mortality has been of the same magnitude in the three groups, and if age constitutes a risk factor for death during the pandemic regardless of level of care. FINDINGS: Individuals living in care homes experienced the highest excess mortality (75- >100% in April, 25-50% in May, 0-25% in June, depending on age). Individuals with home care showed the second highest magnitude (30-60% in April, 15-40% in May, 0-25% in June), while individuals in independent living experienced excess primarily at the highest ages (5-50% in April, 5-50% in May, 0-25% in June). Although mortality rates increased, the age-pattern of mortality during the pandemic resembled the age-pattern observed in previous years. INTERPRETATION: We found stepwise elevated excess mortality by level of care during the first wave of the COVID-19 pandemic in Sweden, suggesting that level of frailty or comorbidities plays a more important role than age for COVID-19 associated deaths. Part of our findings are likely attributable to differences in exposure to the virus between individuals receiving formal care and those living independently, and not only different case fatality between the groups. Although age is a strong predictor for mortality, the relative effect of age on mortality was no different during the pandemic than before. We believe this is an important contribution to the discussion of the pandemic, its consequences, and which groups need the most protection. FUNDING: This study was funded by the Swedish Research Council for Health, Working Life and Welfare (FORTE: grant 2016-07115).

Prenatal ultrasound exposure and children's school performance at age 15-16: follow-up of a randomized controlled trial.

OBJECTIVE: To evaluate the association between prenatal ultrasound exposure and school performance at 15-16 years of age. METHODS: The study population consisted of children born to women who participated in a randomized controlled trial on the second-trimester ultrasound examination in Sweden from 1985 to 1987. Information about the children's grades when graduating from primary school and information on socioeconomic factors was obtained from Swedish nationwide registers. Comparisons were made using linear and logistic regression analyses according to randomization to ultrasound, ultrasound exposure in the second trimester and ultrasound exposure at any time during pregnancy. Boys and girls were analyzed separately. RESULTS: Of the 4756 singleton children from the randomized trial, we identified 4458 (94%) in the National School Register. There were no statistically significant differences in school performance for boys or girls according to randomization or exposure to ultrasound in the second trimester. Compared to those who were unexposed, boys exposed to ultrasound at least once at any time during fetal life had a tendency towards lower mean school grades in general (-4.39 points; 95% CI, -9.59 to 0.81 (max possible, 320) points) and in physical education (-0.45 points; 95% CI, -0.91 to 0.01 (max possible, 20) points), but the differences did not reach significance. CONCLUSION: In general, routine ultrasound examination in the second trimester had no effect on overall school performance in teenagers.

Risk of hypertension amongst Swedish male snuff users: a prospective study.

BACKGROUND: The scientific evidence on whether long-term use of snuff is associated with high blood pressure is limited, inconsistent and based only on cross-sectional data. OBJECTIVE: We aimed at studying the risk of hypertension in relation to long-term use of snuff based on longitudinal data. DESIGN: Repeated health check-ups were offered to all employees in the Swedish construction industry between 1978 and 1993. Blood pressure was measured at the health check-up and information on tobacco use and other risk factors was collected through questionnaires. SETTING: In total, 120 930 never smoking men with information on blood pressure and snuff use at baseline were included. The association of high blood pressure and snuff use at baseline was estimated by logistic regression. Further, 42 055 men were identified as normotensive at baseline and had at least one subsequent health check-up. Through repeated blood pressure measurements and linkage to the Swedish National Inpatient Register, information on hypertension was obtained. Relative risk estimates were derived from Cox proportional hazards regression model. RESULTS: Almost 30% of all men had used snuff. The adjusted odds ratio of high blood pressure amongst snuff users at baseline was 1.23 (95% CI 1.15-1.33) compared to never snuff users. The relative risk of high blood pressure during follow-up was 1.39 (95% CI 1.08-1.79) amongst snuff users and 1.36 (95% CI 1.07-1.72) for hypertension as recorded in the Inpatient Register. CONCLUSION: Use of Swedish moist snuff appears to be associated with a moderately increased risk of hypertension.

alpha-fetoprotein levels in maternal serum during pregnancy and maternal breast cancer incidence.

BACKGROUND: A full-term pregnancy is associated with a reduced risk of breast cancer, but the underlying biologic mechanism has not been elucidated. During pregnancy, maternal serum levels of alpha-fetoprotein, an estradiol-binding protein, rise sharply. In culture, alpha-fetoprotein inhibits the growth of estrogen-sensitive cells, including estrogen-sensitive breast cancer cells. Thus, we investigated whether a high level of alpha-fetoprotein in maternal serum during pregnancy is associated with a reduced risk of breast cancer. METHODS: From a population-based cohort of 42057 pregnant women in Denmark, enrolled in an alpha-fetoprotein-screening program from 1978 through 1996, we obtained a complete reproductive history, vital status, and a possible diagnosis of breast cancer (in 117 women) to the end of follow-up on September 1, 1998. RESULTS: During pregnancy, women with an alpha-fetoprotein level greater than or equal to the median value had a 41% lower risk of breast cancer than women with an alpha-fetoprotein level below the median value (relative risk [RR] = 0.59; 95% confidence interval [CI] = 0.41-0. 85). RRs for breast cancer by mother's age at childbirth were as follows: 29 years or younger, RR = 0.21 (95% CI = 0.08-0.56); 30-34 years, RR = 0.61 (95% CI = 0.32-1.14); 35-37 years, RR = 0.96 (95% CI = 0.49-1.89); and 38 years or older, RR = 0.71 (95% CI = 0.29-1. 75) (P for trend =.02). Further analyses suggested that high levels of alpha-fetoprotein were associated with a reduced incidence of aggressive disease. The most striking finding was that women with high levels of serum alpha-fetoprotein, compared with women with low levels of serum alpha-fetoprotein, showed a particularly reduced incidence of large tumors (>2 cm; RR = 0.24 [95% CI = 0.11-0.50]). CONCLUSION: A high level of alpha-fetoprotein in maternal serum during any pregnancy is associated with a low overall incidence of breast cancer and, in particular, with a low incidence of advanced breast cancer at diagnosis. This association appears particularly strong for a pregnancy occurring at a young age.

Increased prevalence of prior malignancies and autoimmune diseases in patients diagnosed with chronic myeloid leukemia.

We recently reported an increased incidence of second malignancies in chronic myeloid leukemia (CML) patients treated with tyrosine kinase inhibitors (TKI). To elucidate whether this increase may be linked, not to TKI but rather to a hereditary or acquired susceptibility to develop cancer, we estimated the prevalence of malignancies, autoimmune disease (AD) and chronic inflammatory disease (CID) in CML patients prior to their CML diagnosis. Nationwide population-based registers were used to identify patients diagnosed with CML in Sweden 2002-2012 and to estimate the prevalence of other malignancies, AD and CID prior to their CML diagnosis. For each patient with CML, five matched controls were selected from the general population. Conditional logistic regression was used to calculate odds ratios (OR). Nine hundred and eighty-four CML patients were assessed, representing more than 45 000 person-years of follow-up. Compared with matched controls, the prevalence of prior malignancies and AD was elevated in CML patients: OR 1.47 (95% confidence interval (CI) 1.20-1.82) and 1.55 (95% CI 1.21-1.98), respectively. No associations were detected between CML and previous CID. An increased prevalence of other malignancies and AD prior to the diagnosis of CML suggest that a hereditary or acquired predisposition to cancer and/or autoimmunity is involved in the pathogenesis of CML.

Childbearing and the risk of Hodgkin's disease.

The causes of Hodgkin's disease remain incompletely known, but a higher incidence in men than in women has prompted an interest in the role of female sex hormones and reproductive history. Available epidemiological data are, however, contradictory. We analyzed possible associations between parity, age at first birth, and the risk of developing Hodgkin's disease by a linkage between the Swedish Cancer Register and a nationwide Fertility Register. Among women born between 1925 and 1972, 917 cases with Hodgkin's disease and concomitant fertility information were identified. For each case patient, five age-matched controls were randomly selected among women in the Fertility Register. Conditional logistic regression was used to estimate odds ratios of Hodgkin's disease associated with a birth. We found a slightly and nonsignificantly reduced risk of Hodgkin's disease in ever-parous compared with nulliparous women. Among parous women, the number of children was unrelated to risk, whereas there was some evidence of an increased risk with late age at first birth in women under age 45 at diagnosis. No clear temporal relations between childbearing and subsequent risk were discernible in any parity or age group. Although uncontrolled confounding might have affected our results, they do not indicate that hormonal or immunological changes associated with childbearing play a role in the development of Hodgkin's disease.

Does age at the last birth affect breast cancer risk?

We studied the relative importance of age at first and age at last full-term pregnancy (FTP) in a large data set of 3950 cases and 11,510 controls, of whom 2263 cases and 7359 controls had two or more FTPs. We found that a 5-year delay in age at first FTP is associated with an odds ratio of 1.17 (95% confidence interval of 1.05-1.29), whereas a 5-year delay in the age of last FTP is associated with an odds ratio of 1.05 (95% confidence interval 0.96-1.14). Misleading results concerning the relative importance of age at first and last FTP may be obtained when parity is only adjusted in broad categories, uniparous women whose FTP is, simultaneously, both the first and the last are not excluded from the analysis, and age at any FTP between the first and last is not controlled for.

Dual effect of parity on breast cancer risk.

This study examined whether breast cancer risk increased for a short period after childbirth, but decreased after a longer period of time. Data from an international case-control study on breast cancer conducted in the 1960s were used to study the modifying effect of age at enrolment on the relationship between parity and breast cancer risk, comparing first uniparous with nulliparous women, and then biparous versus uniparous women. The statistical analysis was performed by modelling through multiple logistic regression, adjusting for study site, age at menarche, menopausal status and obesity index. Comparing uniparous with nulliparous women, an early age at birth seems to be protective for all periods after birth, whereas a late age at birth imparts a higher risk than nulliparity in the period immediately after birth, which declines with the passage of time. The modification effect by age was not apparent when biparous women with different age at second birth were compared with uniparous women. The results support the hypothesis that pregnancy oestrogens impart a transient increase of maternal breast cancer risk when the full-term pregnancy occurs late in a woman's life.

Incidence of appendicitis during pregnancy.

BACKGROUND: The aetiology and pathogenesis of appendicitis remains unknown. A relation with female sex hormones has been proposed because of a lower incidence among women and incidence variations during the menstrual cycle, but studies have given inconsistent results. Pregnancy constitutes a period of dramatic increases in levels of female sex hormones, but the incidence of appendicitis during childbearing is not known. METHODS: Case-control study of pregnancy status at the time of appendectomy of 53 058 women and of 53 058 population-based age-matched controls. Cases and controls were identified by linkage of the Swedish Inpatient Register and the nation-wide census. Pregnancy status at the time of operation was obtained by linkage with the Swedish Fertility Register. Differences in pregnancy status were analysed using conditional logistic regression and expressed as odds ratios (OR) with 95% CI. RESULTS: Fewer patients than expected with appendicitis were pregnant compared with the controls, especially in the third trimester (OR = 0.49, 95% CI : 0.30-0.79 for perforated and OR = 0.33, 95% CI : 0.28-0.39 for non-perforated appendicitis). CONCLUSIONS: The reduced incidence of appendicitis suggests a protective effect of pregnancy, especially in the third trimester.

Rationales for attending or not attending mammography screening--a focus group study among women in Sweden.

Quantitative findings regarding factors that affect attendance in population-based outreach mammography programmes need to be complemented by descriptive data. The aim of this study was to obtain descriptive insights into the meanings that Swedish women attach to mammograms and their rationales for attending or not attending mammography screening. For this purpose a total of 31 women were recruited to eight focus group discussions. Through inductive analysis, six main themes were identified as important issues in reasoning about attendance or non-attendance in mammography screening: negative experiences, perceived risk factors, knowledge of one's own body, perceived problems with mammography, political, ideological and moral reasoning, and involuntary non-attendance, due to the inability of the screening programme to cover some women (e.g. those with breast implants). In conclusion, there is a need to reinforce the information that mammography is an examination aiming at detecting lumps at an asymptomatic stage. However, the nature and amount of information the women themselves desire is inconsistent. Furthermore, the personal encounter between the individual woman and the staff seems to be of particular importance. An improved dialogue appears an attractive way of adapting the screening situation to meet the varied needs and expectations of the women who are invited.

Predictors of non-attendance in a population-based mammography screening programme; socio-demographic factors and aspects of health behaviour.

The aim of this study was to identify predictors of non-attendance in a population-based mammography-screening programme in central Sweden, on the basis of telephone interviews with 434 non-attendees and 515 attendees identified in a mammography register. Non-attendance was studied in relation to sociodemographic factors, indicators of general health behaviour, self-rated health and experience of cancer in others and own cancer or breast problems. Being single or being non-employed were the only important socio-demographic predictors of non-attendance. Non-attendance was more likely among women who never visited a dentist, had not visited a doctor in 5 years, had never used oral contraceptives or hormone replacement therapy, had never had cervical smear tests, never drank alcohol, smoked regularly, reported no breast cancer in family or friends or own breast problems. We conclude that socio-demographic factors alone do not appear to constitute strong predictors of non-attendance. General health behaviour and previous experience of cancer and breast disease seem to be more important factors. Our results suggest that in the setting of population-based outreach mammography programmes, previous contacts with the health care system and encouragement from health professionals represent determinants of attendance.

Malignant melanoma: reduced risk associated with early childbearing and multiparity.

Pigmentary changes during pregnancy and sex-specific differences in incidence patterns of cutaneous malignant melanoma (CMM) suggest that sex hormones may be involved in the development of CMM. We explored possible associations between childbearing and the risk of CMM in a case-control study "nested' in a nation-wide cohort. A total of 4,779 incident cases of CMM in women aged 24-65 were compared with 23,888 individually age-matched controls. Delayed childbearing was associated with an increased risk of CMM, corresponding to approximately 16% per 5 years. Parous women had a significantly lower risk of CMM compared with nulliparous women; in univariate analysis there was an 8% reduction in risk for each additional birth (odds ratio = 0.92; 95% confidence interval = 0.89-0.95). In multivariate analyses the risk of CMM was best explained by a model including both age at first birth and parity. Age at first birth was the most important variable. Time since most recent birth was unrelated to risk of CMM. These findings indicate that early childbearing and multiparity reduce the risk of CMM. Conceivable explanations are hormonal changes induced by childbearing, enhanced immunologic activity via exposure to fetal antigens during pregnancy, or long-lasting effects of pregnancy-associated hyperpigmentation. Our results need confirmation in studies with proper adjustment for confounding; less sun exposure in young mothers and high parity women may represent an alternative explanation but is unlikely to explain entirely the twofold difference in risk found between extreme categories of age at first birth and parity.

Outcome after lung cancer surgery. Factors predicting early mortality and major morbidity.

OBJECTIVE: This study was undertaken to assess mortality, complications and major morbidity during the first 30 days after lung cancer surgery and to estimate the significance of presurgical risk factors. METHODS: The study was based on all patients referred for surgery for primary lung cancer from 1 January 1987 to 1 September 1999. There were in total 616 patients with primary lung cancer. Three-hundred and ninety-four were men and 222 women. Postoperative events studied were divided into major and minor complications or death during the first 30 days after surgery. The significance of risk factors for an adverse outcome (defined as death or major complication in the first 30 days postoperatively) was assessed by uni- and multivariate logistic regression analyses. RESULTS: During the study period an increasing number of women and of patients older than 70 years underwent surgery. Overall 30-day mortality was 2.9, 0.6% after single lobectomy and 5.7% after pneumonectomy. Major complications occurred in 54 patients (8.8%). Fifty-eight patients (9.5%) had an adverse outcome during the first 30 days. Male gender, smoker, FEV(1)< or =70% of expected value, squamous cell carcinoma and pneumonectomy were risk factors predicting adverse outcome in the univariate model. Pneumonectomy and FEV(1)< or =70%, were the only independently significant factors for adverse outcome. Only pneumonectomy was independently associated with an increased risk for early death. CONCLUSION: Our results show low mortality and morbidity after lung cancer surgery. However, patients with reduced lung capacity and those undergoing pneumonectomy should be treated with great care, as they run a considerable risk of major complications or death during the first 30 days postoperatively. Older age (>70 years) does not appear to be a contraindication to lung cancer surgery, but patients in this group should undergo careful preoperative evaluation.

The shift to and from daylight savings time and motor vehicle crashes.

The objective of the study was to examine whether the shifts to and from daylight savings time in Sweden have short-term effects on the incidence of traffic crashes. A database maintained by the Swedish National Road Administration was used to examine crashes from 1984 through 1995, that occurred on state roads the Monday preceding, the Monday immediately after (index Monday), and the Monday 1 week after the change to daylight savings time in the spring and for the corresponding three Mondays in the autumn. The Mondays 1 week before and after the time changes were taken as representing the expected incidence of crashes. Crash incidence was calculated per 1000 person-years using population estimates for each year of the study. The association between 1 h of possible sleep loss and crash incidence was estimated by the incidence rate ratio from negative binomial regression. The incidence rate ratio was 1.04 (95% CI, 0.92-1.16) for a Monday on which drivers were expected to have had 1 h less sleep, compared with other Mondays. In the spring, the incidence rate ratio for crashes was 1.11 (95% CI, 0.93-1.31) for Mondays after the time change compared to other spring Mondays. The corresponding rate ratio for the fall was 0.98 (95% CI, 0.84-1.15) It was concluded that the shift to and from daylight savings time did not have measurable important immediate effects on crash incidence in Sweden.