Incidence of prostate cancer diagnosis in the eras before and after serum prostate-specific antigen testing

OBJECTIVE

To estimate the incidence of prostate cancer in Olmsted County, Minnesota, from 1983 through 1992 to describe the secular changes that have occurred since the introduction of serum prostate-specific antigen (PSA) testing to the community medical practice in 1987.

DESIGN

Population-based, descriptive epidemiological study with ecological and individual level comparisons over time.

STUDY SETTING

Olmsted County, Minnesota, where the Rochester Epidemiology Project provides passive surveillance of the population for health outcomes.

SUBJECTS

All 511 biopsy-proven incident cases of adenocarcinoma of the prostate diagnosed from 1983 through 1992. The community inpatient and outpatient medical records of all incident cases were reviewed to evaluate the presenting characteristics of men at the time of diagnosis.

RESULTS

The age-adjusted incidence of biopsy-proven prostate cancer increased from 64 per 100,000 person-years in 1983 to 216 per 100,000 person-years in 1992. The increase occurred primarily between 1987 and 1988 and was predominately for organ-confined tumors. The age-specific incidence increased dramatically in this same period among men aged 50 years and older. Among men aged 70 years and older, however, prostate carcinoma incidence rates declined after 1990 following the initial increase. This decline among older men contrasted with community-based estimates of PSA utilization rates, which demonstrated consistent increases since 1987 to nearly 50% of the older population in 1992.

CONCLUSION

These results support the premise that the recent increase in prostate cancer is due in part to the increased utilization of serum PSA testing. Further, the increased incidence appears to be a transient phenomenon due to the depletion of previously undiagnosed cases from the previous pool. Finally, these data suggest that, in terms of stage at diagnosis, early detection efforts may be effective in identifying more early stage (smaller) cancers.

Association between family history of benign prostatic hyperplasia and urinary symptoms: results of a population-based study

Baseline measurements for a population-based prospective cohort study were used to assess the association between family history of enlarged prostate and urinary symptoms. Between December 1989 and March 1991, a group of randomly selected men aged 40-79 years from Olmsted County, Minnesota, was administered a previously validated questionnaire that included questions with wording close to that of the American Urological Association's Symptom Index. A detailed family history of an enlarged prostate was obtained by personal interview, and peak urinary flow rates were measured for each participant. Of the 2,119 men, 440 (21 percent) reported a family history of an enlarged prostate. The age-adjusted odds of having moderate or severe urinary symptoms were elevated among those with a family history relative to those without (odds ratio = 1.3, 95 percent confidence interval 1.1-1.7). With simultaneous control for effects of age and worry about urologic function, the odds ratio remained at 1.3 (95 percent confidence interval 1.0-1.6). Furthermore, this risk was greater for men with relatives diagnosed at a younger age (odds ratio = 2.5, 95 percent confidence interval 1.5-4.3). Men with a family history were also 1.3 times as likely to have an impaired peak urinary flow rate. These findings suggest that men with a family history of an enlarged prostate may be at increased risk for development of symptoms and signs suggestive of benign prostatic hyperplasia and that this risk is greater in men with relatives diagnosed at a younger age. Recognition of this association may help to target early interventions and may lead to further clues about the causes of benign prostatic hyperplasia.

Sexual function of men ages 40 to 79 years: the Olmsted County Study of Urinary Symptoms and Health Status Among Men

OBJECTIVES

Knowledge of male sexual function is somewhat limited because of a lack of current population-based data. This study provides information on sexual function and satisfaction in a population-based sample of men.

METHODS

Men aged 40 to 79 years (n = 2115) were selected randomly from the Olmsted County population for the baseline component of a prospective cohort study (the Olmsted County Study of Urinary Symptoms and Health Status Among Men) during 1989-1990. The men completed a self-administered questionnaire that included questions about sexual concerns, performance, satisfaction, drive, and erectile dysfunction.

RESULTS

For all five sexual parameters queried, the prevalence of problems and dysfunction increased with age. A comparison of men aged 70 to 79 years with men aged 40 to 49 years suggested that older men were more worried about sexual function (46.6% vs 24.9%), had worsened performance compared with a year ago (30.1% vs 10.4%), expressed extreme dissatisfaction with sexual performance (10.7% vs 1.7%), had absent sexual drive (25.9% vs 0.6%), and reported complete erectile dysfunction when sexually stimulated (27.4% vs 0.3%). Logistic regression analyses suggested that sexual dissatisfaction was significantly associated with erectile dysfunction, decreased libido, and the interaction between erectile dysfunction and libido, but not age.

CONCLUSIONS

These population-based cross-sectional data corroborate the previously reported age-related decrease in sexual function. The age-related increase in dissatisfaction could, however, be accounted for primarily by the age-related increase in erectile dysfunction, decreased libido, and the interaction between erectile dysfunction and decreased libido.

Natural history of prostatism: relationship among symptoms, prostate volume and peak urinary flow rate

We describe relationships among symptoms, prostate volume and peak urinary flow rate in an age stratified, community based random sample of white men 40 to 79 years old with no prior prostate surgery, prostate cancer or other conditions known to interfere with voiding. Symptoms were assessed with an instrument comparable to the American Urological Association symptom index. Prostate volume was estimated by transrectal ultrasonography and peak urinary flow rate was measured by a portable device. Subject age was significantly associated with symptom score but accounted for only 3% of its variation, while prostate volume and peak urinary flow rate explained only an additional 10% of the symptom variability. The odds (95% confidence interval) of moderate to severe symptoms increased with age from 1.9 (1.1 to 3.1), 2.9 (1.7 to 5.0) and 3.4 (1.8 to 6.1) for men 50 to 59, 60 to 69 and 70 to 79 years old, respectively, relative to men 40 to 49 years old. Adjusting for age, the odds of moderate to severe symptoms were 3.5 times greater for men with prostatic enlargement (more than 50 ml.) than for men with smaller prostates, while the odds were similarly increased (2.4-fold) for men not achieving a peak urinary flow rate of 10 ml. per second. Estimated odds changed little when other cutoff points were considered for peak urinary flow rate (15 ml. per second) or prostate volume (40 ml.). These results, based on randomly selected white men, suggest a somewhat stronger, albeit modest, relationship among symptoms, prostate size and urinary flow rate than previously reported in clinic based studies. The strength of these relationships is comparable to that found with other diseases.

Exploration and exploitation of clinical databases

Clinical data repositories represent a potential gold mine of information and knowledge. Rapid access to such information can help bridge the gap between clinical care and research, support clinical and executive decision making, and improve the quality of care. A clinical database can be used in four ways: to display information about an individual patient (results reporting); to find data on a patient with similarities to one being seen (case finding); to describe a group of patients with at least one attribute in common (cohort description); and to analyze data patterns in terms of trends or relationships (predictive modeling). It seems unlikely that many important clinical questions will be subject to randomized clinical trials because of the ethics, logistics, and expense that would be involved. Evolving statistical and epidemiological methods allow us to approach these clinical data repositories with the purpose of building predictive models, but a clear understanding of the limitations of routinely collected clinical data and the inherent biases is necessary. The largest barrier to using routinely collected clinical data is not the limitations of the data themselves, but rather the lack of a data paradigm for the decision-maker. We present some of the problems and pitfalls in obtaining and using routinely collected data, based upon the use of ClinQuery at Boston's Beth Israel Hospital and the resources and traditions at the Mayo Clinic.

The role of community-based longitudinal studies in evaluating treatment effects. Example: benign prostatic hyperplasia

To determine the appropriateness and effectiveness of medical interventions, it is necessary to understand both the natural history of the conditions the interventions are intended to treat or prevent, and the normal reference ranges of tests used in diagnosis and management. The acquisition of this information through clinic-based studies can yield misleading conclusions owing to selection bias. What appears to be intervention-related variation in outcomes may be variation in the magnitude and form of selection bias. To minimize selection bias, population-based studies are required. Results from the Olmsted County Study of Urinary Symptoms and Health Status Among Men were used to show how a population-based longitudinal study of the natural history of benign prostatic hyperplasia (BPH) can complement research programs of the BPH Patient Outcomes Research Team. Population-based studies of disease natural history are a necessary part of medical outcomes research, and deserve greater emphasis in the medical treatment effectiveness research initiative of the Agency for Health Care Policy and Research.

An overview of statistical methods for the classification and retrieval of patient events

Statistical methods that can support text retrieval are becoming an increasing focus of medical informatics activities. We overview our adaptation of existing knowledge sources to create pseudo-documents for concept based latent semantic indexing. Experience demonstrated this tack of limited practical value, since retrieval performance was invariably unsatisfactory. We discovered this was due in part to the introduction of a vocabulary gap between the queries and the cases we sought to retrieve. In part to address this problem, and to avail our large body of humanly coded text as a knowledge source, we developed a least squares fit alternative for the computer assisted indexing and retrieval of biomedical texts. This technique demonstrates equivalent or superior retrieval performance when compared to all other textual retrieval techniques. It does not depend upon elaborate knowledge bases, lexicons, or thesauri. It is a promising technique for classifying and retrieving the large volumes of clinical text.

Sampling strategies in a statistical approach to clinical classification

This paper studies the sampling strategies for the Expert Network (EexNet), a statistical learning system used for patient record classification at the Mayo Clinic. The goal is to achieve high accuracy classification at an affordable computational cost in very large applications. The learning curves of ExpNet were observed with respect to the choice of training resources, the size, vocabulary coverage and category coverage of a training set, and the category distribution over training instances. A method combining advantages of different sampling strategies is proposed and evaluated using a large training corpus. As a result, Expert Network has achieved its nearly-optimal classification accuracy (measured by average precision) using a relatively small training set, with a fast real-time response which satisfies the needs of human-machine interaction.

Medical information retrieval and WWW browsers at Mayo

Medical information retrieval from "Master Sheet" entries specially indexed for research retrieval has been part of the Mayo culture since 1909. Providing easy to use and universally available WWW access to these and other patient information databases at Mayo via browsers, shines a bright light on issues of privacy and confidentiality, user authentication, need to know, data transmission security, and technical details of interfacing disparate databases on a spectrum of platforms to many types of workstations using a variety of browsers. We review our recent experience, and generalize pertinent issues.

Do prostate size and urinary flow rates predict health care-seeking behavior for urinary symptoms in men?

OBJECTIVES

To estimate the association between health care-seeking behavior for urinary dysfunction and clinical, physiologic, and anatomic measures of disease.

METHODS

A randomly selected sample (n = 475) of men aged 40 to 79 years from Olmsted County, Minnesota, was administered a previously validated questionnaire that assessed the frequency of and bother associated with urinary symptoms and health care-seeking behavior in the past year. Peak urinary flow rates were measured with a standard urometer and prostatic volume was determined by transrectal ultrasound.

RESULTS

Overall, 21 of the 475 men (4%) had seen a doctor in the past year for urinary symptoms. Men with moderate to severe symptoms (American Urological Association [AUA] Symptom Scores > 7) were 3.4 times as likely (95% confidence interval [CI] = 1.4, 8.3) to have sought medical care in the past year as men with none to mild symptoms. Men with enlarged prostates (> 40 mL) were 3.9 times as likely to have sought health care (95% CI = 1.6, 9.6), whereas men with depressed peak urine flow rates (< 10 mL/s) were only slightly more likely to have sought health care for urinary symptoms (odds ratio = 2.1, 95% CI = 0.7, 6.5). Overall, 76% of men who had sought medical care had prostatic enlargement, depressed peak urine flow rates, or moderate-severe symptoms (sensitivity). In contrast, only 55% of men who did not seek health care for urinary symptoms in the past year had mild symptoms, normal prostatic volume, and normal peak urine flow rates (specificity).

CONCLUSIONS

These data suggest that clinical, physiologic, and anatomic measures of prostatism do not adequately distinguish the men who seek medical care for their urinary symptoms from those who do not. There remain some factor(s) that apparently lead some men with minor disease to seek care and that prevent men with measurable disease from seeking care.

The natural history of prostatism: the effects of non-response bias

BACKGROUND

In epidemiological studies, non-response may raise the question of generalizability to the target population. Most investigations have not been able to access data that could provide information about the potential impact of non-response bias.

METHODS

A 55% response rate was realized at baseline for a prospective cohort investigation of the natural history of benign prostatic hyperplasia in Olmsted County, Minnesota, during 1989-1991 (the Olmsted County Study of Urinary Symptoms and Health Status Among Men). This prompted a preliminary study of potential non-response bias among full participants, partial participants and complete non-responders. The medical diagnostic index maintained by the Rochester Epidemiology Project was used to ascertain the prevalence of specific conditions in the 9 years prior to study inception.

RESULTS

The age-adjusted period prevalence rate for benign prostatic hyperplasia (%) was 9.6 (95% confidence interval [CI]: 8.1-11.0) for full participants, 8.2 (95% CI: 5.8-10.6) for partial participants and 5.3 (95% CI: 3.6-6.9) for complete non-responders. Other urologic diagnoses followed the same pattern. However, age-adjusted prevalence rates for general medical examination history and major non-urologic morbidities were decidedly similar across response groups.

CONCLUSIONS

These data suggest response may have been driven, in part, by concerns about urologic disease. However, the similarity in non-urologic diagnoses and general medical examinations provide some preliminary reassurance that the 55% response rate did not necessarily compromise generalizability.

Obesity and benign prostatic hyperplasia

Abdominal obesity increases the estrogen-to-androgen ratio and may increase sympathetic nervous activity, both hypothesized to influence the development of benign prostatic hyperplasia and the severity of urinary obstructive symptoms. In 1986 and 1987, men aged 40-75 years who were participants in the Health Professionals Follow-up Study and who were without prior diagnosis of cancer or prostatectomy provided data on weight, height, and waist and hip circumferences. The men were followed for incidence of prostatectomy for benign prostatic hyperplasia up to January 1992. In addition, the frequency and severity of symptoms of urinary obstruction were assessed among respondents to a questionnaire in 1992. Among 25,892 men who provided complete information for both surgery and symptoms, 837 men had surgery for benign prostatic hyperplasia, and 2,581 of those without surgery reported frequent urinary symptoms. After adjustment for age, smoking, and body mass index, abdominal obesity was related to prostatectomy (odds ratio (OR) = 2.38, 95% confidence interval (CI) 1.42-3.99, for those with a waist circumference > or = 43 inches (109 cm) relative to those with a waist circumference < 35 inches (89 cm); p trend < 0.0001) and with frequent urinary symptoms among those without prostatectomy (OR = 2.00, 95% CI 1.47-2.72; p < 0.0001). Body mass index, hip circumference, and waist-to-hip ratio were not associated with benign prostatic hyperplasia independently of waist circumference. These results suggest that abdominal obesity in men may increase the frequency and severity of urinary obstructive symptoms and may increase the likelihood that such obese men will undergo a prostatectomy.

Lower-extremity motor neuropathy associated with surgery performed on patients in a lithotomy position

BACKGROUND

Motor neuropathy of a lower extremity is well-recognized as a potential complication of procedures performed on patients in a lithotomy position. Most of this awareness is based on anecdotal reports, however, and the incidence and risk factors for this complication have not been reported.

METHODS

We retrospectively reviewed the perioperative courses of 198,461 consecutive patients who underwent 1 of 56 surgical procedures historically performed on patients in a lithotomy position at the Mayo Clinic, Rochester, Minnesota, from 1957 to 1991 inclusive. The medical diagnoses of patients who had procedures in a lithotomy position were scanned for 26 diagnoses associated with neuropathy. Persistent neuropathy of the lower extremity was defined as a motor deficit of at least 3 months' duration. Risk factors anecdotally associated with persistent neuropathy were analyzed by comparing identified cases of neuropathy to controls in a 1:3 case-control study.

RESULTS

Persistent neuropathies after procedures performed on patients in a lithotomy position were identified in 55 cases for a rate of 1 per 3,608. Multivariate risk factors for development of a persistent neuropathy of a lower extremity included duration in lithotomy of 4 h or longer, a body mass index (kilograms per squared meter) of 20 or less, and a history of smoking within 30 days of the procedure. Regional anesthetic techniques were not found to be associated with an increased risk of neuropathy. Of the 53 patients who lived at least 1 yr after their procedure, 24 (45%) required either prosthetic or ambulatory support for persistent foot drop or leg weakness.

CONCLUSIONS

These data suggest that prolonged duration in lithotomy and patient risk factors, including very thin body habitus and smoking in the preoperative period, are associated with the development of a lower-extremity neuropathy after procedures performed on patients in a lithotomy position. A reduction of time in the lithotomy position may be particularly worthwhile for patients with these risk factors.

Effect of several recruitment strategies on response rates at baseline in a prospective cohort investigation. The Olmsted County Study of Urinary Symptoms and Health Status among Men

Epidemiologic survey response rates were studied in relation to maneuvers introduced to improve acceptance: (a) variation in invitation letters, (b) the use of a brochure with the recruitment mailing, and (c) options for interview location. The baseline population-based survey of a prospective cohort investigation of the natural history of benign prostatic hyperplasia was used. Invitations to participate were mailed to eligible, randomly selected men aged 40 to 79 years from the Olmsted County, Minnesota, population during 1989 to 1991. Of the 3874 men identified, 2119 (55%) participated. Overall, there was no difference in response rate according to invitation characteristics (chi 2(5) = 8.02, P = 0.16). Nevertheless, response rates varied with age (chi 2(7) = 30.9, P < 0.001) and home location (rural versus Rochester city; chi 2(1) = 76.9, P < 0.001). This suggests the innovations used to bolster acceptance did not materially improve response rates. Further, since response rates were highest for men aged 60 to 74 years, men with more symptoms and free time may have joined the cohort more often than others.

Cigarette smoking and prostatism: a biphasic association?

OBJECTIVE

To evaluate the association between cigarette smoking and prostatism in a community-based setting using standardized urinary symptom scores, peak urinary flow rates, and prostatic volume as indicators of disease.

METHODS

A population-based cohort of 2,115 Caucasian men aged forty to seventy-nine years from Olmsted County, Minnesota, was administered a previously validated questionnaire that elicited information on frequency of urinary symptoms (approximating the American Urological Association's symptom index), and a detailed history on cigarette smoking, including both amount and pack-years of smoking. Peak urinary flow rates were measured by a standard uroflowmeter (Dantec 1000). The prostatic volume was measured for a subsample of 471 men by transrectal ultrasound.

RESULTS

Compared to never-smokers, smokers were less likely to have moderate to severe urinary symptoms (age-adjusted odds ratio 0.82; 95% confidence interval [CI] 0.61 to 1.08). This varied by smoking intensity, however; in men who smoked less than 1 pack a day the age-adjusted odds ratio was 0.53 (95% CI 0.33 to 0.83) and among men smoking 1 to 1.4 packs a day, the odds ratio was 0.87 (95% CI 0.56 to 1.36). For men who smoked 1.5 packs or more a day, the odds ratio was elevated at 1.32 (95% CI 0.84 to 2.07). Smokers were less likely to have peak flow rates less than 15 mL/sec compared with never-smokers (age- and voided volume-adjusted odds ratio 0.48; 95% CI 0.35 to 0.66), or prostatic volume greater than 40 mL (odds ratio 0.54; 95% CI 0.19 to 1.55).

CONCLUSIONS

These data from a community-based sample suggest that light or moderate smokers are less likely to have moderate to severe prostatism, whereas heavy smokers are at least as likely to have moderate to severe prostatism compared with never-smokers.

Natural history of prostatism: worry and embarrassment from urinary symptoms and health care-seeking behavior

OBJECTIVE

To assess the interrelationships among psychosocial symptoms of worry and embarrassment about urinary function, prevalent urinary symptoms, psychological well-being, and health care-seeking behavior in a population-based cohort of men.

METHODS

A cohort of 2,119 men aged forty to seventy-nine years, randomly selected from the Olmsted County, Minnesota population between December 1989 and March 1991, were administered a previously validated questionnaire that elicited information about the frequency of urinary symptoms, the degree to which they were perceived as a bother, and if the participant had seen a doctor in the previous twelve months for evaluation of any of these urinary symptoms. Psychological well-being was assessed by a subset of the Psychological General Well-Being Index, and sociodemographic information was also sought.

RESULTS

Urinary symptom indices (measured by American Urological Association frequency and bother scores and psychological general well-being subscales) were significantly associated with worry and embarrassment about urinary symptoms in bivariate analyses. Multiple logistic regression analyses demonstrated that men with moderate or severe urinary symptoms or impaired psychological well-being were more likely to be worried or embarrassed about their urinary symptoms than men with mild symptoms. Furthermore, men who were worried about their urinary function were more likely to have sought medical care for their symptoms than men who were not worried. The association between health care-seeking behavior and embarrassment was especially strong among men with little bother associated with their urinary symptoms.

CONCLUSIONS

Worry and embarrassment about urinary symptoms reflect quality-of-life issues that appear important in the health care-seeking behavior of men with prostatism. The results underscore findings that prevalent urinary symptoms alone do not determine a man's health care-seeking behavior, and treatment for psychosocial symptoms may be beneficial in some men with symptoms of prostatism.

Prostate cancer in Rochester, Minnesota (USA), from 1935 to 1989: increases in incidence related to more complete ascertainment

Prostate cancer incidence among White men in the United States climbed steadily from 45 per 10(5) person-years (PY) during 1945-54 to 102 per 10(5) PYs in 1988. To determine whether this increase might be the result of changing diagnostic practices, we examined trends in incidence and method of diagnosis in Rochester, Minnesota (US), from 1935 to 1989. We found a parallel increase in Rochester in non-autopsy diagnoses from 44 (95 percent confidence interval [CI] = 29-58) cases per 10(5) PYs in 1935-44 to 71 (CI = 52-89) cases per 10(5) PYs in 1985-87 which was driven by diagnoses prompted by digital rectal examination. There was no evidence that an increasing proportion of cases was found as the result of procedures to treat the symptoms of benign prostatic hyperplasia. Including autopsy diagnoses, incidence was stable over this extended interval and was 77 per 10(5) PYs (CI = 58-97) in 1935-44 and 72 per 10(5) PYs (CI = 53-91) in 1985-87. Incidence more than doubled after introduction of diagnostic serum prostate-specific antigen (PSA) assay and was 179 per 10(5) PYs (CI = 145-214) in 1988-89. We conclude that prostate-cancer incidence rates are influenced strongly by diagnostic practices and that national increases could reflect, to a large extent, more complete and earlier ascertainment rather than more frequent disease.

Correlation between urinary flow rate, voided volume, and patient age in a community-based population

This investigation is unique in that is the first study to use a large number of randomly chosen men from the community to establish the relationship between: 1) urinary flow rate and voided volume, 2) urinary flow rate and patient age, 3) voided volume and patient age, and 4) the time to achieve peak urinary flow rate and patient age. The results clearly indicate that both the peak and mean urinary flow rate decreases with advancing age. The voided volume also diminishes with increasing age. The time required to achieve the peak urinary flow rate, however, appears to be independent of the patient's age. Irrespective of age, approximately 9 seconds are required to achieve the maximum urinary flow rate. Because the urinary flow rate is dependent upon both the quantity of urine voided and the patient's age, it is necessary to account for the influence of both parameters when establishing the "normal range" for peak urinary flow rate. To this end, nomograms estimating the peak urinary flow rate percentile as a function of voided volume and patient age have been generated and are included in this report. These nomograms, based on a large, randomly selected population, should make peak urinary flow rate a more reliable diagnostic modality for assessing bladder outlet obstruction. They also should be most useful to the practicing clinician when interpreting the results of a peak urinary flow rate determination for a patient presenting with symptoms of prostatism.

Medical data and knowledge management by integrated medical workstations: summary and recommendations

The health care professional workstation will function as an interface between the user and the patient data as well as an interface pertinent medical knowledge. Appropriate knowledge focus will require the workstation to recognize the concepts and structure of patient data, and understand the scope and access methods of knowledge sources. Issues are organized around five major themes: (i) structure, (ii) reliability and validation, (iii) views, (iv) location, and (v) ethical and legal. Conventional database representations can effectively address data structure and format variations that will inevitably persist in local data stores. The reliability of data and the validation of knowledge are critical issues that may determine the ultimate utility of clinical workstations. Alternative views of patient information and knowledge sources represent the true power of an intelligent data portal, represented by a well-designed clinical workstation. Both data and knowledge are optimally represented in decentralized information networks, although the confidentiality and ownership of this information must be respected. Evolutionary progress toward consistent representations of knowledge and patient data will be facilitated by the establishment of self-documentation standards for the developers of data encoding systems and knowledge sources, perhaps extended from the preliminary model afforded by the Unified Medical Language System (UMLS).

An application of Expert Network to clinical classification and MEDLINE indexing

An effective and efficient learning method, Expert Network (ExpNet), is introduced in this paper. ExpNet predicts the related categories of an arbitrary text based on a search of its nearest neighbors in a set of training texts, and a reasoning from the expert-assigned categories of these neighbors. Evaluations in patient-record text classification and MEDLINE document indexing show a performance of ExpNet in recall and precision comparable to the Linear Least Squares Fit (LLSF) mapping method, and significantly better than other methods tested. We also observed that ExpNet is much more efficient than LLSF in computation. The total training and testing time on the patient-record text collection (6134 texts) was 4 minutes for ExpNet versus 96 minutes for LLSF; on the MEDLINE document collection (2344 documents), the total time was 15 minutes for ExpNet versus 4.6 hours for LLSF. It is evident in this study that human knowledge of text categorization can be statistically learned without expensive computation, and that ExpNet is such a solution.

An evaluation of computer assisted clinical classification algorithms

The Mayo Clinic has a long tradition of indexing patient records in high resolution and volume. Several algorithms have been developed which promise to help human coders in the classification process. We evaluate variations on code browsers and free text indexing systems with respect to their speed and error rates in our production environment. The more sophisticated indexing systems save measurable time in the coding process, but suffer from incompleteness which requires a back-up system or human verification. Expert Network does the best job of rank ordering clinical text, potentially enabling the creation of thresholds for the pass through of computer coded data without human review.

Natural history of prostatism: factors associated with discordance between frequency and bother of urinary symptoms

The objective of this study was to assess the association between frequency and bother of urinary symptoms in a population-based cohort of men and to identify psychosocial factors that are related to reporting heightened or subdued bother relative to symptom frequency. The survey was conducted among men aged forty to seventy-nine years in Olmsted County, Minnesota, the baseline component of a prospective cohort study. Men were queried about the frequency of urinary symptom occurrence and the perceived bother associated with the symptoms. A regression analysis of American Urologic Association (AUA) bother scores on AUA frequency scores demonstrated a tight correspondence (r2 = 0.71). Men with bother scores greater than predicted from their frequency scores were more likely to have sought health care for their urinary symptoms than men whose bother was close to predicted (14 versus 5 percent, respectively). These men with heightened bother were older, poorer, more anxious, and had lower general psychologic well-being scores than the men whose bother was similar to that expected from their reported frequency. Men whose bother was lower than would be expected were less likely to have sought health care for urinary symptoms in the past year (3%) but were of similar age and socioeconomic status as compared with men whose bother was close to expected. These men, however, were more depressed than men whose bother was commensurate with reported frequency. While the men who reported greater bother than expected from their symptom frequency were more likely to have sought medical care for urinary symptoms in the past year, it is not clear whether this greater health-care-seeking behavior is because bother captures an additional component of urologic disease or is a manifestation of psychosocial differences.

Similar levels of urological symptoms have similar impact on Scottish and American men--although Scots report less symptoms

Urinary symptoms and the extent to which they interfere with living activities were compared in 2 community-based investigations that enrolled men 40 to 79 years old who were randomly sampled from Olmsted County, Minnesota (2,119) and the Forth Valley of Scotland (1,385). Both investigations included symptom questions with wording that is nearly identical to that of the American Urological Association (AUA) symptom index. Following AUA scoring conventions we grouped scores into mild (AUA score 0 to 7), moderate (score 8 to 19) and severe (score 20+) categories. Minnesota men had symptoms that were more frequent, more bothersome and caused greater interference with living activities than did Scottish men of comparable age (p < 0.0002). However, within each symptom score category, the extent to which symptoms interfered with living activities was essentially the same in both populations. Although there appear to be important differences in urinary symptom prevalence between Scotland and Minnesota, the AUA symptom index provides a consistent measure of the extent to which urinary symptoms interfere with living activities in both populations. These findings support use of the AUA symptom index in the diagnostic evaluation of men with benign prostatic hyperplasia.

A prospective study of dietary fat and risk of prostate cancer

BACKGROUND

The strong correlation between national consumption of fat and national rate of mortality from prostate cancer has raised the hypothesis that dietary fat increases the risk of this malignancy. Case-control and cohort studies have not consistently supported this hypothesis.

PURPOSE

We examined prospectively the relationship between prostate cancer and dietary fat, including specific fatty acids and dietary sources of fat. We examined the relationship of fat consumption to the incidence of advanced prostate cancer (stages C, D, or fatal cases) and to the total incidence of prostate cancer.

METHODS

We used data from the Health Professionals Follow-up Study, which is a prospective cohort of 51529 U.S. men, aged 40 through 75, who completed a validated food-frequency questionnaire in 1986. We sent follow-up questionnaires to the entire cohort in 1988 and 1990 to document new cases of a variety of diseases and to update exposure information. As of January 31, 1990, 300 new cases of prostate cancer, including 126 advanced cases, were documented in 47855 participants initially free of diagnosed cancer. The Mantel-Haenszel summary estimator was used to adjust for age and other potentially confounding variables. Multiple logistic regression was used to estimate relative risks (RRs) when controlling simultaneously for more than two covariates.

RESULTS

Total fat consumption was directly related to risk of advanced prostate cancer (age- and energy-adjusted RR = 1.79, with 95% confidence interval [CI] = 1.04-3.07, for high versus low quintile of intake; P [trend] = .06). This association was due primarily to animal fat (RR = 1.63; 95% CI = 0.95-2.78; P [trend] = .08), but not vegetable fat. Red meat represented the food group with the strongest positive association with advanced cancer (RR = 2.64; 95% CI = 1.21-5.77; P = .02). Fat from dairy products (with the exception of butter) or fish was unrelated to risk. Saturated fat, monounsaturated fat, and alpha-linolenic acid, but not linoleic acid, were associated with advanced prostate cancer risk; only the association with alpha-linolenic acid persisted when saturated fat, monounsaturated fat, linoleic acid, and alpha-linolenic acid were modeled simultaneously (multivariate RR = 3.43; 95% CI = 1.67-7.04; P [trend] = .002).

CONCLUSION

The results support the hypothesis that animal fat, especially fat from red meat, is associated with an elevated risk of advanced prostate cancer.

IMPLICATIONS

These findings support recommendations to lower intake of meat to reduce the risk of prostate cancer. The potential roles of carcinogens formed in cooking animal fat and of alpha-linolenic acid in the progression of prostate cancer need to be explored.