Use of patient-reported global assessment measures in clinical trials of chronic pain treatments: ACTTION systematic review and considerations.

ABSTRACT: Establishing clinically meaningful changes in pain experiences remains important for clinical trials of chronic pain treatments. Regulatory guidance and pain measurement initiatives have recommended including patient-reported global assessment measures (eg, Patient-Global Impression of Change [PGIC]) to aid interpretation of within-patient differences in domain-specific clinical trial outcomes (eg, pain intensity). The objectives of this systematic review were to determine the frequency of global assessment measures inclusion, types of measures, domains assessed, number and types of response options, and how measures were analyzed. Of 4172 abstracts screened across 6 pain specialty journals, we reviewed 96 clinical trials of chronic pain treatments. Fifty-two (54.2%) studies included a global assessment measure. The PGIC was most common (n = 28; 53.8%), with relatively infrequent use of other measures. The majority of studies that used a global assessment measure (n = 31; 59.6%) assessed change or improvement in an unspecified domain. Others assessed overall condition severity (n = 9; 17.3%), satisfaction (n = 8; 15.4%), or overall health status/recovery (n = 5; 9.6%). The number, range, and type of response options were variable and frequently not reported. Response options and reference periods even differed within the PGIC. Global assessment measures were most commonly analyzed as continuous variables (n = 24; 46.2%) or as dichotomous variables with positive categories combined to calculate the proportion of participants with a positive response to treatment (n = 18; 34.6%). This review highlights the substantial work necessary to clarify measurement and use of patient global assessment in chronic pain trials and provides short- and long-term considerations for measure selection, reporting and analysis, and measure development.

Mammography screening and increased incidence of breast cancer in Wisconsin.

The age-adjusted incidence of breast cancer among Wisconsin women age 40 and older has increased by almost one third since the early 1980s. To estimate what portion of the observed increase in breast cancer incidence is due to mammography screening, we developed a model of the lead-time effect of this screening test and its impact on incidence. The model incorporates annual age-specific information including 1) the expected number of cases, 2) the rate of screening, 3) the detection ratio of screening mammography, and 4) the lead time of screening mammography. For women 40 years of age and older, the model predicts a 25% increase in incidence, compared with an observed increase of 28%. Overall, mammography screening explains 74% of the difference between the expected and observed number of cases over the study period. A greater portion of the increase in incidence among postmenopausal women is attributed to mammography screening than among younger women. The increase in the use of mammography appears to account for most but not all of the increased incidence of breast cancer in Wisconsin.

Trends in the surgical treatment of prostate cancer in Wisconsin, 1989-1991.

BACKGROUND: Radical prostatectomy (removal of the prostate gland and seminal vesicles) is usually considered a definitive treatment for localized prostate cancer. Although a sharp increase and wide geographic variation in radical prostatectomy rates have been recently documented, the reasons for this increase and the factors that make men diagnosed with the disease more likely to be treated surgically are not well known. PURPOSE: Our purpose was to examine trends in the use of surgical treatment for prostate cancer, as well as the factors associated with the choice of surgical treatment and how these factors changed in Wisconsin in the period 1989 through 1991. METHODS: We carried out a population-based cohort study. A cohort of Wisconsin men diagnosed from 1989 through 1991 with prostate cancer was identified through the Wisconsin cancer reporting system. To determine which men diagnosed with prostate cancer were treated with surgery, we linked prostate cancer records to the Wisconsin hospital discharge database. The outcome measured was radical prostatectomy within 6 months from the date of the diagnosis of prostate cancer. RESULTS: The yearly number of prostate cancer cases reported from 1989 through 1991 rose 33%, from 2468 to 3278. During the same period, the yearly number of radical prostatectomies rose 226%, from 283 to 922. Patients diagnosed in 1991 were twice as likely to have surgery as those diagnosed in 1989, the proportion of cases receiving surgical treatment rising from 12% to 25%. Patients who were white, less than 65 years of age, had a cancer reported to be at regional stage, and who were first reported by large hospitals were more likely to be treated surgically. CONCLUSIONS: The use of surgery to treat prostate cancer has increased dramatically in Wisconsin, doubling in a 3-year period, despite the fact that studies have not shown surgery to be more effective than other options for many patients. The increase observed in the rate of surgery was about the same across age, race, stage at diagnosis, and hospital size. The reasons for this increase are probably multifactorial. IMPLICATIONS: Risks, costs, and benefits of surgical treatment should be carefully compared with those of alternative management approaches. Patients should be enabled to make an informed decision, based on the current best evidence, on the treatment option they prefer.

Smoking and the risk of peptic ulcer disease among women in the United States.

Although multiple studies support a causal relationship between smoking and peptic ulcers in men, data for women are limited. Therefore, we used data from the First National Health and Nutrition Examination Survey Epidemiologic Follow-up Study, a nationally representative prospective study of US adults, to evaluate the impact of smoking on the incidence of peptic ulcers in women. The study cohort included 2851 women who had not been diagnosed as having a peptic ulcer prior to the baseline interview. Among these women, 140 (4.9%) developed peptic ulcer disease. During 12.5 years of follow-up, the estimated cumulative incidence of ulcers was 10.0% for current smokers, 6.4% for former smokers, and 5.4% for never smokers. After adjusting for age, education, regular aspirin use, coffee consumption, and use of alcohol, current smokers were 1.8 times more likely to develop ulcers than never smokers (95% confidence interval, 1.2 to 2.6); the risk of peptic ulcer increased as the amount smoked increased. During the time of this study, we estimate that approximately 20% of incident peptic ulcer cases among US women were attributable to cigarette smoking.

The 10-year incidence of overweight and major weight gain in US adults.

We estimated the 10-year incidence of major weight gain (a gain in body mass index of greater than or equal to 5 kg/m2 and overweight (a body mass index of greater than or equal to 27.8 for men and greater than or equal to 27.3 for women) in US adults using data from the First National Health and Nutrition Examination Survey Epidemiologic Follow-up Study. Persons aged 25 to 74 years at baseline were reweighed a decade after their initial examination (men, 3727; women, 6135). The incidence of major weight gain was twice as high in women and was highest in persons aged 25 to 34 years (men, 3.9%; women, 8.4%). Initially overweight women aged 25 to 44 years had the highest incidence of major weight gain of any subgroup (14.2%). For person not overweight at baseline (men, 2760; women, 4295), the incidence of becoming overweight was similar in both sexes and was highest in those aged 35 to 44 years (men, 16.3%; women, 13.5%). We conclude that obesity prevention should begin among adults in their early 20s and that special emphasis is needed for young women who are already overweight.

Self-perceived stress and the risk of peptic ulcer disease. A longitudinal study of US adults.

BACKGROUND: Although many physicians and laypersons believe that stress plays a role in the occurrence of peptic ulcer disease, the importance of stress in the pathogenesis of peptic ulcers remains controversial. METHODS: To investigate the relationship between perceived stress and peptic ulcer disease we used data from the National Health and Nutrition Examination Survey Epidemiologic Follow-up Study--a nationally representative cohort study of US adults. This analysis included 4511 persons who had not previously been diagnosed with peptic ulcer disease. RESULTS: At baseline, 68% of the cohort perceived themselves as stressed. During 13 years of follow-up, 208 persons developed ulcers; the cumulative incidence of ulcers was 7.2% for persons who were stressed and 4.0% for persons who were not. After we adjusted for age, sex, education, smoking status, and regular aspirin use, persons who perceived themselves as stressed were 1.8 times more likely to develop ulcers than those who did not (95% confidence interval, 1.3 to 2.5). We also found a graded relationship between the perceived amount of stress and the incidence of peptic ulcers; relative to nonstressed persons, the relative risk of developing an ulcer was 1.4, 1.9, 2.3, 2.4, and 2.9 at five increasing levels of stress. CONCLUSIONS: These findings suggest that persons who perceive their lives as stressful may be at increased risk for the development of peptic ulcer disease.

Evaluation of reactions to food additives: the aspartame experience.

Despite the widespread use of chemical food additives, few criteria exist to evaluate consumer reports of adverse reactions. We analyzed 231 consumer complaints associated with the food additive aspartame. We developed a methodologic approach to evaluate all complaints by adapting general criteria used to investigate adverse reactions to medications. Complaints were ranked according to the effects of cessation and rechallenge. Using this method, we found no clear symptom complex that suggests a widespread public health hazard associated with aspartame use; however, we identified some case reports in which the symptoms may be attributable to aspartame in commonly-consumed amounts. The systematic application of pre-defined review criteria, such as those described here, to monitor consumer complaints related to food additives will help identify products that warrant more focused clinical studies.

Nonsteroidal anti-inflammatory drug use and protection against colorectal cancer in women.

Several epidemiological studies have identified an association between nonsteroidal anti-inflammatory drug (NSAID) use and colorectal cancer risk in women. We examined this association in a population-based case-control study in Wisconsin women. Between 1991 and 1992, 184 women ages 40-74 years with colorectal cancer were identified through the statewide cancer registry and 293 population-based control women were randomly selected via telephone. Regular NSAID use was defined as at least twice weekly for 12 months or longer. After adjusting the data for age, controls were more likely than cases to report regular NSAID use (38 versus 27%). Following adjustment for age, prior sigmoidoscopy use, family history of large bowel cancer, and body mass index, women who regularly used NSAIDs were approximately one-third less likely to be diagnosed with colorectal cancer compared to women who did not use NSAIDs [odds ratio (OR), 0.65; 95% confidence interval (CI), 0.40-1.03]. A statistically significant effect of duration of use was identified, although the ORs did not show a consistent trend. No significant effect of frequency of NSAID use was observed. When the type of NSAID used was examined (aspirin or nonaspirin), subjects who used nonaspirin compounds had a statistically significantly lower risk of colorectal cancer (OR, 0.43; 95% CI, 0.20-0.89), compared to nonusers, whereas aspirin users had only a small, nonsignificant reduction in cancer risk (OR, 0.79; 95% CI, 0.46-1.36). These data add support to the hypothesis that regular NSAID use is associated with lower colorectal cancer risk in women and suggest that the type of NSAID used may be important.

Risk factors for carcinoma in situ of the breast.

As more women obtain screening mammograms regularly and at younger ages, the diagnosis of breast carcinoma in situ becomes more frequent. To examine whether risk factors for carcinoma in situ correspond with risk factors for invasive breast cancer, we analyzed data from a population-based case-control study conducted in 1988-1990. We identified newly diagnosed cases of carcinoma in situ (n = 301) and invasive breast cancer (n = 3789) in women 18-74 years of age from Wisconsin's statewide tumor registry. Cases and population controls (n = 3999) completed structured telephone interviews. Overall, associations with risk of carcinoma in situ in relation to many reproductive life-style risk factors were similar to those associated with risk of invasive disease. Women who reported a family history of breast cancer had a 2-fold elevated risk of carcinoma in situ (odds ratio, 2.67; 95% confidence interval, 2.00-3.57). Personal history of benign biopsied breast disease also increased risk of carcinoma in situ (odds ratio, 2.19; 95% confidence interval, 1.62-2.95). Subgroup analysis suggested that high vitamin A intake and high alcohol intake may be associated with risk of ductal but not lobular carcinoma in situ. These data support the presence of common risk factors between in situ and invasive breast cancer.

Early-life physical activity and postmenopausal breast cancer: effect of body size and weight change.

It is not yet known whether early-life physical activity reduces the risk of developing breast cancer. Subgroup analyses according to menopausal status and body mass may help clarify this association. Data from a population-based case-control study of female residents of Wisconsin, Massachusetts, Maine, and New Hampshire were used to examine associations between body mass and breast cancer risk. Cases (n = 4614) were identified by each state's tumor registry; controls (n = 5817) were randomly selected from population lists. Frequency of participation in strenuous physical activity when 14-22 years of age, weight at age 18 and 5 years before interview, height, and other factors were ascertained through structured telephone interviews. Adjusted odds ratios (ORs) and 95% confidence intervals (CIs) were computed using logistic regression. Reductions in postmenopausal breast cancer risk associated with strenuous physical activity were greatest for women in the fourth quartile of body mass index at age 18; the OR for women with the highest activity frequency on average (> or =once/day) was 0.45 (95% CI = 0.26-0.79). Associations with frequency of activity also varied by weight change. Compared to women with no activity and little adult weight gain, frequent physical activity was associated with reduced postmenopausal breast cancer risk in women who had lost weight since age 18 (OR = 0.19, 95% CI = 0.05-0.70) or had gained little or modest amounts of weight (weight gain: first tertile, OR = 0.36, 95% CI = 0.05-0.85; second tertile, OR = 0.31, 95% CI = 0.14-0.66). Weighted MET score analyses yielded similar but less inverse results. These findings suggest that the reduced risk of postmenopausal breast cancer associated with frequent, early-life physical activity may be greatest in women who, over the adult years, either lost weight or gained only modest amounts.

Decreasing trends in Reye syndrome and aspirin use in Michigan, 1979 to 1984.

The incidence of Reye syndrome has been decreasing in Michigan, perhaps as a result of decreased aspirin use among children. To evaluate possible changes in the frequency of aspirin use, 199 families in Tecumseh, MI, with children younger than 18 years of age were interviewed by telephone in February 1981 and again in February 1983. Based on the reported use of medications for colds or influenza between 1981 and 1983, fewer parents gave aspirin (56% v 25%), but acetaminophen use did not change (59% v 55%). Younger parents and parents who had heard of the association between aspirin and Reye syndrome were more likely to stop giving aspirin. More parents chose to use either no medication or medications containing neither aspirin nor acetaminophen (6% v 32%) for the treatment of colds or influenza. Approximately 90% of parents who chose not to give aspirin for fever also gave medications for colds or influenza that did not contain aspirin. These results suggest that fewer children are receiving aspirin during illnesses that may precede Reye syndrome. The associated decrease in the incidence of Reye syndrome tends to support the hypothesis that the use of aspirin increases the risk for the development of Reye syndrome.

Epidemiological model and cost-effectiveness analysis of tuberculosis treatment programmes in Indonesia.

An epidemiological model of tuberculosis, based on the natural history of tuberculosis and the control programmes in Indonesia, was constructed. This model was used for estimating future tuberculosis-prevented cases and costs for three treatment strategies--the 100% standard course, the 100% short course, and the existing strategy (a combination of 65% standard course and 35% short course)--in accordance with the master plan of the Indonesian Government's tuberculosis control programme. A cost-effectiveness analysis of the three strategies confirmed that the short-course strategy was the most cost-effective. Sensitivity analysis, which applied a broad range of parameters, continued to confirm the short-course strategy as the most cost-effective. If the short-course strategy had been applied in 1980 instead of the existing strategy (using the most likely parameters), the short-course strategy would prevent 1.8 million sputum-positive cases and would save 61.0 million dollars by the year 2000.

Hepatitis B vaccination programs for hospital workers: results of a statewide survey.

To assess the implementation of hepatitis B virus (HBV) vaccination programs for hospital workers, we mailed questionnaires to all 229 licensed Michigan hospitals. The response rate was 96% (221/229); of these, 68% (150/221) had vaccination programs. Although multiple hospital characteristics were associated with the presence of a vaccination program, characteristics that independently predicted the presence of a program were medical school affiliation, nonpsychiatric specialty, and the existence of a hepatitis B immune globulin protocol. The most common reason given (56%, 40/71) for the absence of a program was insufficient worker risk of hepatitis B infection; this response was frequent in psychiatric (91%, 10/11) and rural hospitals (61%, 11/18). Among high-risk workers, attending physicians were less likely than other high-risk workers to be included in vaccination programs (68% vs. 95%, respectively). Fear of vaccine-associated acquired immunodeficiency syndrome was most frequently cited as the primary reason for vaccine refusal. We conclude that unwarranted fears about the vaccine's safety need to be dispelled, that high-risk physicians should be included in vaccination programs, and that rural and psychiatric hospital policies reflect their perceived risk of occupational HBV infection.

Trends in United States ovarian cancer mortality, 1979-1995.

OBJECTIVE: To describe the epidemiology of ovarian cancer mortality in the United States from 1979 to 1995. METHODS: The mortality data of the Centers for Disease Control and Prevention were accessed using the Wide-ranging Online Data for Epidemiologic Research (WONDER). We selected all deaths among women with International Classification of Diseases, Ninth Revision (ICD-9) code 183.0 (ovarian malignant neoplasm). Mortality data for the years 1979-1995 were age-adjusted to the United States 1990 female population, and mortality rates for each year were calculated for females of all ages by age category, by race, and by geographic location. Trends were obtained for the periods 1979-1983 to 1991-1995, and the impact on the number of ovarian cancer deaths was calculated. RESULTS: Age-adjusted ovarian cancer mortality rates have changed little in the United States from 1979 to 1995, but rates are increasing in older women (65 years and older) and decreasing in younger women. Age-adjusted mortality rates are higher among whites than in blacks. Ovarian cancer mortality rates are higher in northern compared with southern states. CONCLUSION: The trends in ovarian cancer mortality among younger and older women parallel published changes in incidence and may be due to changes in risk factors, such as the use of oral contraceptives. The reasons for the higher ovarian cancer death rates in northern states are unknown. Better understanding of how modifiable risk factors and treatment methods affect ovarian cancer mortality trends is needed.

Cervical cancer screening among women with and without hysterectomies.

OBJECTIVE: To compare the rate of Papanicolaou testing in a population-based sample of women with medical documentation of 1) total hysterectomy for benign conditions, 2) total hysterectomy for malignant conditions, and 3) hysterectomy with cervix intact to rates among women who had not had a hysterectomy. METHODS: The Marshfield Epidemiologic Study Area was used to identify a retrospective cohort of women with hysterectomies age-matched to women without hysterectomies. This study compares the Papanicolaou test rate per year (outcome) by hysterectomy status (exposure) for women with total hysterectomy for benign reasons (n=197), total hysterectomy for malignancy (n=75), supracervical hysterectomy (n=43), and no hysterectomy (n=315). RESULTS: Compared with women who did not have a hysterectomy (nonexposed), women with a hysterectomy (exposed) for benign reasons had significantly fewer Papanicolaou tests; on average, one less test every 3 years (mean difference=-0.34 tests/year, P < .001). Contrary to this, women with a malignancy-related hysterectomy had significantly more tests than their nonexposed counterparts (mean difference=0.87 tests/year, P < .001); nearly one additional test per year. Finally, women with supracervical hysterectomies had the same rate of testing as their nonexposed counterparts (mean difference=-0.03 tests/year, P=.62); on average, one test every 2.5 years. CONCLUSION: This study demonstrates that Papanicolaou testing rates vary by type and reason for hysterectomy. Women with hysterectomies for benign reasons may be receiving from two to three times as many tests as needed. Notably, women with intact cervices following hysterectomy have similar testing rates (one every 2.5 years) as women without hysterectomies. This has direct implications for leaving a woman's cervix intact given normal cytology at the time of hysterectomy.

The Wisconsin Firearm-Related Injury Surveillance System.

CONTEXT: Firearm-related injuries rank second only to motor vehicle-related injuries as a cause of injury death in Wisconsin. OBJECTIVE: To evaluate the attributes of the Wisconsin Firearm-Related Injury Surveillance System. DESIGN: A structured surveillance system evaluation using predetermined criteria. SETTING: A passive surveillance system linking administrative data from existing state-funded inpatient hospitalization and mortality databases. PARTICIPANTS: State health department. MAIN OUTCOME MEASURES: Attributes assessed included simplicity, flexibility, acceptability, sensitivity, predictive value positive, representativeness, timeliness, resources, and data quality. RESULTS: The use of two existing state databases simplifies data acquisition and linkage. However, hospital discharge and vital records databases are not sufficiently flexible to collect perpetrator and circumstance information. Acceptability is high because of state-mandated reporting to both databases. For firearm-related injuries requiring hospitalization, the system's predictive value positive is 97% when E codes are compared with data from chart reviews. The system is considered timely because annual data from the hospital discharge and vital records systems can be obtained, linked, analyzed, and reported by September of the subsequent year. The system is sustainable largely because existing software is used for annual evaluations, which requires less than 2 weeks of staff time. CONCLUSIONS: With minimal resources and time, the Wisconsin Firearm-Related Injury Surveillance System uses existing state government databases to describe and report the burden of firearm-related injuries. Additional information on circumstances, perpetrators, and weapons involved are available but additional resources are needed to integrate this information with existing data.

Screening mammography for women 50 years of age and older: practices and trends, 1987.

Recently, public and private efforts have been mounted to promote screening mammography. To assess recent trends in the percentage of women 50 years of age and older who have had a screening mammogram, we analyzed data from interviews from women from 33 states who participated in the 1987 Behavioral Risk Factor Surveillance System. Our study group included 8,402 women 50 years of age and older who had visited a physician for a routine checkup in the last year; among these 8,402 women, only 29% reported having had a screening mammogram in the past year. However, of the women in the study group, the percentage who had a screening mammogram in the last year showed a relative increase of 38% during 1987, from 24% for women interviewed in the first quarter of 1987 to 33% for women interviewed in the fourth quarter. However, not all groups of patients benefited equally from the observed trend--the absolute and relative increases in the percentage of women screened were lowest for women who were older, less educated, in low-income groups, and who had poor personal health practices. Although the percentage of women 50 years of age and older who reported being screened increased dramatically during 1987, special efforts are needed to reach the patient groups that are being left behind in the trend toward increased use of screening mammograms.

Patterns of self-reported drinking and driving in Michigan.

We analyzed a statewide telephone survey of Michigan adults to determine patterns of self-reported drinking and driving. The estimated prevalence of drinking and driving was 13.5 percent for men and 2.9 percent for women; the highest prevalence was among 18- to 24-year-old men (32 percent). Based on these estimates, over half a million Michigan adults drank and drove on over one million occasions during the month preceding the survey. Most drinking drivers (93 percent) reported binge drinking, yet 70 percent of them otherwise reported only moderate routine alcohol consumption, that is, they consumed fewer than 14 drinks a week on average. Because we were concerned about the validity of self-reports, we compared the patterns of self-reported alcohol use with the patterns of alcohol-related motor vehicle crashes. The pattern of self-reported drinking and driving (using age-, sex-, and region-specific estimates) was highly correlated with the pattern of injury in alcohol-related crashes (r = .96; p less than .0001). Self-reported patterns of alcohol use may be used to identify persons at highest risk for being injured or dying in a motor vehicle crash.

Dietary and weight control practices among persons with hypertension: findings from the 1986 Behavioral Risk Factor Surveys.

The National Heart, Lung, and Blood Institute has concluded that reducing sodium intake, controlling weight, and moderating alcohol consumption can help patients with hypertension control their blood pressure. To determine whether such patients have adopted recommended dietary practices, we analyzed data collected in 1986 from telephone surveys of adults in 26 states (no. = 34,395). The self-reported dietary practices that we evaluated were: use of table salt, alcohol consumption, and weight control practices. In comparison with persons who did not have hypertension (no. = 26,848), those with hypertension who were receiving pharmacological therapy ("treated hypertensives," no. = 5,025) were more likely to report limiting their use of table salt (odds ratio [OR] = 2.5) and were less likely to have their weight controlled (OR = 0.4). Although overweight persons with hypertension were more likely than persons with normal blood pressure to attempt to lose weight, most have not included exercise in their weight loss efforts. There was no difference between persons who do not have hypertension and treated patients with hypertension in their use of alcohol. Untreated persons with hypertension (no. = 2,378) were less likely to limit their use of table salt and less likely to moderate their use of alcohol than treated persons with hypertension but otherwise reported similar dietary practices. Dietetic practitioners may need to emphasize moderation of alcohol use and the use of physical activity to lose weight when counseling patients with hypertension.

A method to guide community planning and evaluation efforts in tobacco control using data on smoking during pregnancy.

BACKGROUND: Effective community based tobacco control programmes are critical for state and nationwide impact. However, there is little discussion in the literature of methods for setting local objectives which use locally collected data and account for historical variation in progress. OBJECTIVES: To develop and illustrate a method that uses locally available birth certificate data to model trends in tobacco use during pregnancy among women giving birth, predict future prevalence, and use predictions to set community specific tobacco control objectives. DATA SOURCE: Vital statistics. Wisconsin standard birth certificates, 1990-2000, which record the smoking status of the mother during pregnancy. DATA ANALYSIS: Trends in the prevalence of smoking during pregnancy in Wisconsin statewide and in all counties (n = 72) were modelled using linear regression of log prevalence on year. Model fit was assessed using R(2). Regression slopes, indicating estimated relative annual percentage change in prevalence, were used to predict prevalence in 2005, and objectives were calculated as a 20% reduction from the predicted prevalence in 2005. CONCLUSIONS: Modelling trends in the prevalence of smoking using locally collected data enables communities to set reasonable future tobacco control objectives that account for historical trends in progress.