Increasing response rates from physicians in oncology research: a structured literature review and data from a recent physician survey.

Although the physician survey has become an important tool for oncology-focused health services research, such surveys often achieve low response rates. This mini-review reports the results of a structured review of the literature relating to increasing response rates for physician surveys, as well as our own experience from a survey of physicians as to their referral practices for suspected haematologic malignancy in the United States. PubMed and PsychINFO databases were used to identify methodological articles assessing factors that influence response rates for physician surveys; the results were tabulated and reviewed for trends. We also analysed the impact of a follow-up telephone call by a physician investigator to initial non-responders in our own mailed physician survey, comparing the characteristics of those who responded before vs after the call. The systematic review suggested that monetary incentives and paper (vs web or email) surveys increase response rates. In our own survey, follow-up telephone calls increased the response rate from 43.7% to 70.5%, with little discernible difference in the characteristics of early vs later responders. We conclude that in addition to monetary incentives and paper surveys, physician-to-physician follow-up telephone calls are an effective method to increase response rates in oncology-focused physician surveys.

Prevention-oriented life styles and diffusion of cholesterol screening and awareness: Massachusetts behavioral risk factor survey, 1987-1991.

Universal screening of the adult population for detection of elevated serum cholesterol has been recommended. We examined the relation of eight risk factors for morbidity and mortality (hypertension, overweight, inactivity, tobacco use, safety belt nonuse, binge alcohol consumption, driving after alcohol consumption, and chronic alcohol consumption) to adoption of cholesterol screening and to awareness of cholesterol level. Data were collected through the Massachusetts Behavioral Risk Factor Surveillance System between 1987 and 1991 (mean number of respondents interviewed annually, 1240). We compared trends in prevalence of cholesterol screening and awareness within risk groups defined on the basis of the presence or absence of each risk factor. Cholesterol screening prevalence increased from 46.8% in 1987 to 67.9% in 1991. Overweight and hypertensive respondents were more likely to have been screened than nonoverweight or normotensive respondents; for the other six risk factors, individuals at increased risk were less likely to have been screened. The difference in cholesterol screening prevalence between increased-risk and lower-risk respondents increased between 1987-1988 and 1990-1991 for four risk factors. Prevalence of awareness of cholesterol level increased from 7.8% in 1987 to 35.4% in 19991. Trends by risk status were comparable to those observed for cholesterol screening. Individuals already motivated toward a preventive life style appear to be those most likely to avail themselves of a new prevention possibility.

Reproducibility of the women's module of the Behavioral Risk Factor Surveillance System questionnaire.

The Behavioral Risk Factor Surveillance System (BRFSS) is designed to provide statewide estimates of the prevalence of preventive health practices, including screening. We assessed the reproducibility of responses to the women's health module, which covers breast and cervical cancer screening, hysterectomy, and pregnancy. A random sample of women in Massachusetts (n = 91; response rate for the repeat interview, 70.0%) and a separate random sample of minority women in the state (n = 179; response rate for the repeat interview, 69.4%) were interviewed by telephone twice, 21 to 94 days apart. Differences across administrations in mean prevalence of screening were small. Concordance exceeded 85% for almost all the variables examined, but tended to be lower for nonwhite respondents. After correction for agreement occurring by chance, moderate to excellent values of kappa (range, 0.41 to 0.86) were observed. The women's health module of the BRFSS questionnaire yields highly consistent group mean estimates of prevalence when administered repeatedly to the same individuals. Individual reproducibility is excellent, but may be reduced among minority respondents.

Reproducibility of responses to telephone interviews: demographic predictors of discordance in risk factor status.

The reproducibility of responses to the Behavioral Risk Factor Surveillance System questionnaire was examined across the demographic strata used by the Centers for Disease Control and Prevention and state health departments for reporting prevalence estimates (specifically age, sex, income, employment, and marital status), as well as race/ethnicity, which has been previously examined. The authors administered the questionnaire twice, 21-94 days apart, to randomly selected residents of Massachusetts (response rates: first administration, 68% of eligible households; second administration, 68% of persons who completed the first interview). Initial interviews were conducted in March and October 1992. Among 448 respondents to both interviews, group mean distributions of seven demographic characteristics and 19 risk factors were highly consistent across the two interviews. Discordance in individual risk factor status ranged from 1.2% to 21.8% (median, 7.8%) and was symmetric in direction, i.e., as many respondents were considered at increased risk on the basis of the first interview and at low risk on the basis of the second interview as the reverse. Kappas ranged from 0.30 to 0.90 (median, 0.75). Education, household income, and interval between administrations were not associated with prevalence of discordance for any risk factor. Sex, age, race/ethnicity, marital status, and employment status were each predictive of variation in discordance for one or more risk factors, but no consistent effect of any individual demographic characteristic across risk factors was observed. The questionnaire has relatively uniform and generally good reproducibility across all demographic strata used for monitoring the Health Objectives for Year 2000 and other chronic disease surveillance activities.

The Behavioral Risk Factor Surveillance System questionnaire: its reliability in a statewide sample.

The reliability of the Behavioral Risk Factor Surveillance System questionnaire was assessed in a random sample of adults (n = 122) and a separate sample of Black and Hispanic adults (n = 200) in Massachusetts. The questionnaire was administered twice, 21 to 44 days apart, by telephone (210 completed reinterviews, 65% response rate for second administration). There were no statistically significant differences in the distribution of demographic or risk factor variables across administrations. Individual-level reliability (kappa for categorical variables, correlation for continuous variables) for demographic characteristics was more than 0.80 for White respondents and more than 0.60 for Black and Hispanic respondents. Employment and income were reported less consistently than other variables. Reliability coefficients for behavioral risk factors were generally above 0.70. Exceptions were variables with extreme distributions. These data support the use of the Behavioral Risk Factor Surveillance System questionnaire for surveillance and research.

Maternal ethnicity and birthweight among blacks.

This paper examines the association of ethnicity and birthweight, adjusted for other maternal and infant characteristics, among black women who gave birth in Massachusetts from 1987 through 1989. Data are drawn from the standard certificate of live birth, which includes questions on race and ethnicity/ancestry as well as birthweight; maternal sociodemographic and biological characteristics; access to prenatal care; and infant characteristics. The study cohort consists of 18,571 black infants and a comparison group of 206,358 non-Hispanic white infants. Infants whose mothers reported their race as black were further categorized into six ethnic groups: American, Haitian, West Indian, Cape Verdean, Hispanic, and other black. In addition to descriptive analyses, we used multiple linear regression to measure the association between ethnicity, other characteristics, and birthweight; and we used multiple logistic regression to measure the odds ratio of low birthweight (ranging from 500 g to 2499 g) for the six black ethnic groups, adjusted for other characteristics. Results indicate that Americans have lower mean birthweight and generally higher levels of risk than other black ethnic groups. Compared to the reference group of non-Hispanic whites, Americans (OR = 1.49), other blacks (OR = 1.41), and West Indians (OR = 1.37) have significantly elevated relative risks of low birthweight.

Predictors of medical care utilization by independently living adults with spinal cord injuries.

A prospective study of 87 independently living adults with spinal cord injury (SCI) as a major disabling condition showed the following average annual health care utilization rates: 1.3 hospital admissions, 16.8 days hospitalized, 1.7 emergency room (ER) visits, and 22.4 outpatient contacts (in person or by telephone). Those hospitalized (n = 66) experienced a mean of 22.2 days hospitalized per person per year. Mean length of stay (LOS) was 11.1 days per admission. Stepwise regression analysis indicated no statistically significant (p less than or equal to .05) predictors of hospital admissions. There were three independent predictors of days hospitalized (greater age, fewer years of education, and more days hospitalized during the previous year), three predictors of days hospitalized for those hospitalized only (greater age, fewer years of education, and longer hospital LOS during the previous year), one predictor of LOS (self-assessment of health), three of emergency room (ER) visits (more unmet instrumental activities of daily living needs, lack of organizational memberships, and more ER visits during the previous year), and five predictors of outpatient contacts (greater age, less satisfaction with health care providers' expressions of concern for their health, lower frequency of leaving apartments, lower levels of life satisfaction, and nonparticipation in a managed medical care demonstration project). Many predictors of health services utilization are immutable. However, changes which facilitate social interaction and changes in the organization of health services may reduce certain types of medical care utilization by people with SCI.

Alcohol, tobacco, and cannabis use by independently living adults with major disabling conditions.

Data from an 18-month prospective study of 205 adults with major disabling conditions revealed that there is a positive but not statistically significant association (p greater than .10) between respondents' scores on a scale which measures both numbers and amounts of substances (alcohol, tobacco, and cannabis) used and bed disability days. There are also nonsignificant positive relationships between scale scores and five of six measures of utilization of medical care.

The epidemiology of medical care utilization by severely-disabled independently-living adults.

A prospective study of the medical care utilization experience of 205 severely-disabled independently-living adults in Eastern Massachusetts shows that there was a mean of 0.83 +/- 1.26 hospital admissions, 9.9 +/- 22.7 hospital days, 1.5 +/- 2.31 emergency room (ER) visits, and 26.88 +/- 44.4 outpatient contacts per person per year. Among those hospitalized, the mean experience was 16.2 +/- 27.1 days per person per year; mean length-of-stay was 9.3 +/- 14.7 days per admission. Regression analysis indicates that those with spinal cord injuries as major disabling conditions were significantly more likely to be hospitalized. So were those with lower self-assessments of health, higher levels of depressions, and more baseline ER visits. Self-assessment of health is a significant predictor of hospital days for the total cohort (including those with no admissions); so are age at onset of disability (greater age; higher risk), and bed disability days in the month before the baseline survey (more disability days; higher risk). Among those hospitalized, the total number of days hospitalized is significantly related to both age at onset of disability (later onset; more days) and baseline days hospitalized (greater number; more days). Lengths-of-stay are significantly related to two factors; age and age at onset of disability (in both cases, greater age associated with longer stays). Prior ER visits are a significant predictor of subsequent ER visits (more baseline; more subsequent); so are respondents' reported satisfaction with their participation in their medical care (lower reported satisfaction; more ER visits), organizational affiliations, and frequencies of contacts with friends or relatives. Higher levels of social interaction (i.e. organizational affiliation and more frequent social contacts) were associated with more ER visits. Prior contacts with physicians, nurse-practitioners, or physician-assistants was the most powerful predictor of subsequent outpatient contacts (more baseline; more subsequent). There were also significant relationships between subsequent contacts and respondents' assessments of their health relative to others with similar disabilities (relatively worse health; more contacts), age (greater age; more contacts), and baseline ER visits (more visits; more contacts).

A prospective evaluation of the effect of managed care on medical care utilization among severely disabled independently living adults.

We conducted an 18-month longitudinal evaluation of a model-managed medical care program for severely disabled, independently living adults. Regression analyses using an additive model (no interaction effects) suggest that persons in the study group did not have statistically significantly different utilization experiences than members of the comparison group. Regression analyses that include interaction effects suggest that, for certain segments of the cohort, the study group's utilization experience was significantly lower than that of members of the comparison group. Persons in the study group with higher baseline emergency room (ER) utilization had significantly fewer hospital admissions (P = 0.0055). The participants with better self-assessments of health experienced significantly fewer hospital days per person (P = 0.0075) and days per person hospitalized (P = 0.0056), and persons with organizational affiliations reported significantly fewer ER visits (P = 0.0264).

Rehospitalization and spinal cord injury: cross-sectional survey of adults living independently.

A cross-sectional survey of 96 people living independently with spinal cord injuries (SCI) in Eastern Massachusetts shows that 57% had been hospitalized at least once in the year before the survey. Sample means were 1.0 admissions and 16.0 days/person/year. Eight percent of the sample (eight persons) accounted for 22% of admissions and 59% of total hospital days. For those hospitalized, the mean was 1.7 admissions and 45.1 days/person/year. Mean length-of-stay was 34.7 days/admission. Multiple regression analysis shows that three variables appear to be independently related to increased numbers of admissions: self-assessment of health; place of residence; and age (younger respondents at higher risk). One variable is independently associated with total days of hospitalization: leaving home at least once daily (as opposed to less frequently) is associated with lower risk. There were no statistically significant relationships between either numbers of hospitalizations or total days hospitalized and ADL or IADL status, education, employment, medical insurance, household composition, gender, age at onset of disability, time since onset of disability, substance use (alcohol, cannabis, or tobacco), level of SCI lesion, or social supports.

Mandatory seatbelt law support and opposition in New England--a survey.

Less than one-fifth of the U.S. population consistently wears automobile seatbelts. Automatic seatbelts or air bags will be required in all new cars, unless States whose total population equals two-thirds of the nation's adopt mandatory seatbelt use laws by 1989, according to the U.S. Department of Transportation. In July 1984, New York State passed the first mandatory seatbelt law in the nation, followed by New Jersey, Illinois, Missouri, Michigan, and New Mexico. A telephone survey of 2,982 randomly selected drivers examined belt use and support for mandatory belt use laws in the six New England States. Drivers with little education and low income, younger drivers, and drivers who drove after heavy drinking or marijuana use, or both, were least likely to wear seatbelts. Beliefs that seatbelts are not effective in reducing injury risk and are uncomfortable were more common among those not wearing belts. Sixty percent of drivers favored a mandatory seatbelt use law. The most important predictors of opposition to a belt use law were beliefs that seatbelts are ineffective, inconvenient, and uncomfortable. Opposition was also more likely among persons who reported that they drove after marijuana use, or heavy drinking, or both, drove more miles per year, exhibited a low frequency of seatbelt use, and perceived a low probability of personal crash involvement.

Employee drinking patterns and accidental injury: a study of four New England states.

The relationship between alcohol use and accidental injury was examined in an anonymous telephone survey of 1740 randomly sampled employed adults (49% men) residing in four New England states in 1982 and 1983. Respondents reported 383 accidents requiring medical attention during the year prior to the interview, 34 involving hospitalization. Of the accidents, 41% occurred at work, 32% during recreational and other activities, 19% at home and 8% in traffic incidents. Respondents were stratified by self-reported levels of alcohol consumption and were compared on accident rates. Drinking at work was also examined for an association with accident involvement. Logistic regression analyses controlled for variability associated with age, sex, marital status, occupation and drug use. Drinking at work was not significantly or independently associated with job accidents but an average daily intake of 5+ drinks elevated the relative risk, compared with abstainers, as follows: 1.7 for any accidental injury, 3.8 for injuries requiring hospitalization and 2.0 for job-based accidental injuries. The relative risks associated with the use of other psychoactive drugs were 1.7 for work accidents and 2.4 for accident-associated hospitalizations. Of the respondents, 13% reported having ever had a drug or alcohol problem and 2% reported a current problem. Fourteen percent reported that there were counselors at their jobs to help with such problems. Of those who had ever experienced a problem, 11% had sought help from outside programs and only 4% had sought help at work.