Participation rates in epidemiologic studies.

A survey's response rate is a conventional proxy for the amount of response bias contained in that study. While there are more theoretical opportunities for bias when response rates are low rather than high, there is no necessary relationship between response rates and bias (O'Neill, Marsden, and Silman 1995; Asch, Jedrziewski, and Christakis 1997). A review of physician surveys published between 1985 and 1995 found an average response rate of 61 percent for all surveys of physicians and an average response rate of 52 percent for surveys with more than 1,000 observations (Cummings, Savitz, and Konrad 2001). There was no significant decline in response rates to published physician surveys across that time, but it is unclear whether that pattern has continued.

By definition, not much information is available about nonrespondents. It has been estimated that only 18 percent of physician survey articles perform any type of comparison between responders and nonresponders with the remainder having only response rates to serve as an indicator of bias (Cummings, Savitz, and Konrad 2001). Direct examinations of response bias for health care professional groups have generally found only minimal amounts of response bias in surveys (Barton et al. 1980; Hovland, Romberg, and Moreland 1980; Locker and Grushka 1988; McCarthy, Koval, and MacDonald 1997; Thomsen 2000; Kellerman and Herold 2001). Surveys of patients and the full population have consistently shown substantially more response bias (Sheikh and Mattingly 1981; Benfante et al. 1989; Brennan and Hoek 1992; Diehr et al. 1992; Vestbo and Rasmussen 1992; Prendergast, Beal, and Williams 1993; Walsh 1994; Blanker et al. 2000; Kotaniemi et al. 2001; Barchielli and Balzi 2002; Fowler et al. 2002; Mazor et al. 2002; Solberg et al. 2002; Partin et al. 2003; Van Loon et al. 2003). Substantial response bias greatly limits the generalizability of survey findings (Asch, Jedrziewski, and Christakis 1997; Cummings, Savitz, and Konrad 2001). Studies focused specifically on assessing response bias for physician surveys are lacking. It is important to determine if response rates for physicians are falling, and if decreasing response rates are associated with increasing response bias.

The American Academy of Pediatrics (AAP) has conducted many surveys of pediatricians through its Periodic Survey of Fellows and other programs. The AAP also maintains an administrative database that contains demographic information about its members. Using this database along with data from the American Board of Pediatrics administrative database, age, gender, membership type, and location data were merged with respondent/nonrespondent information for 50 surveys of pediatricians. The objectives in doing this were (1) to monitor response rates across time, (2) to examine systematic response bias for several different pediatrician characteristics, and (3) to explore response bias as a function of survey response rates.

Response rates and response bias for 50 surveys of pediatricians.

Thoracic spine pain (TSP) is experienced across the lifespan by healthy individuals and is a common presentation in primary healthcare clinical practice. However, the epidemiological characteristics of TSP are not well documented compared to neck and low back pain. A rigorous evaluation of the prevalence, incidence, correlates and risk factors needs to be undertaken in order for epidemiologic data to be meaningfully used to develop evidence-based prevention and treatment recommendations for TSP. A systematic review method was followed to report the evidence describing prevalence, incidence, associated factors and risk factors for TSP among the general population. Nine electronic databases were systematically searched to identify studies that reported either prevalence, incidence, associated factors (cross-sectional study) or risk factors (prospective study) for TSP in healthy children, adolescents or adults. Studies were evaluated for level of evidence and method quality. Of the 1389 studies identified in the literature, 33 met the inclusion criteria for this systematic review. The mean (SD) quality score (out of 15) for the included studies was 10.5 (2.0). TSP prevalence data ranged from 4.0–72.0% (point), 0.5–51.4% (7-day), 1.4–34.8% (1-month), 4.8–7.0% (3-month), 3.5–34.8% (1-year) and 15.6–19.5% (lifetime). TSP prevalence varied according to the operational definition of TSP. Prevalence for any TSP ranged from 0.5–23.0%, 15.8–34.8%, 15.0–27.5% and 12.0–31.2% for 7-day, 1-month, 1-year and lifetime periods, respectively. TSP associated with backpack use varied from 6.0–72.0% and 22.9–51.4% for point and 7-day periods, respectively. TSP interfering with school or leisure ranged from 3.5–9.7% for 1-year prevalence. Generally, studies reported a higher prevalence for TSP in child and adolescent populations, and particularly for females. The 1 month, 6 month, 1 year and 25 year incidences were 0–0.9%, 10.3%, 3.8–35.3% and 9.8% respectively. TSP was significantly associated with: concurrent musculoskeletal pain; growth and physical; lifestyle and social; backpack; postural; psychological; and environmental factors. Risk factors identified for TSP in adolescents included age (being older) and poorer mental health. TSP is a common condition in the general population. While there is some evidence for biopsychosocial associations it is limited and further prospectively designed research is required to inform prevention and management strategies.

/pdf/thoracic-spine-pain-in-the-general-population-prevalence-1k8043b7rn.pdf

Thoracic spine pain in the general population: prevalence, incidence and associated factors in children, adolescents and adults. A systematic review.

Background: Most national-level mammography data are for ever-had and most-recent screening. There are few national-level data on rates of repeat, on-schedule mammography, and on the prevalence and correlates of repeat mammography. It is also important to investigate the similarity of correlates for repeat and recent mammography.

Methods: Analyses were from data for women ages 45 to 75 in the 2002 to 2003 Health Information and National Trends Survey (HINTS 2003; N = 1,581). The two dependent variables were self-report of repeat mammography (two exams on schedule, based on an every-other-year interval) and recent mammography only (one mammogram within the past 2 years).

Results: The prevalence of recent mammography was 81.6% (95% confidence interval, 79.1-84.1) and for repeat mammography was 72.2% (95% confidence interval, 69.0-75.4). An access to care variable combining insurance coverage and regular source of care was the strongest sociodemographic correlate of both mammography indicators. Most other sociodemographic variables were not associated with mammography status. Five psychosocial/behavioral variables were associated with both mammography indicators (smoking status, attention to health information, knowledge of screening interval, worry about breast cancer, and recent mood status). Correlates were very similar for repeat and recent mammography.

Conclusions: Although access to care had the strongest association with mammography, psychosocial and behavioral variables did better as a group than the sociodemographic variables. A standard set of such variables should be considered for all national surveys. (Cancer Epidemiol Biomarkers Prev 2006;15(11):2093–101)

https://cebp.aacrjournals.org/content/cebp/15/11/2093.full.pdf

Correlates of repeat and recent mammography for women ages 45 to 75 in the 2002 to 2003 health information national trends survey (HINTS 2003)

INTRODUCTION Missing abstinence outcomes are a universal challenge for tobacco cessation studies. The biases inherent in the complete case analysis, where cases missing outcomes are omitted, are well known. Sophisticated statistical methodologies are available to address these biases but are not widely applied in tobacco cessation trials. Within tobacco cessation research, a widely used strategy is penalized imputation (PI), wherein cases missing cessation outcomes are assigned a "currently using tobacco" status. STATISTICAL BASIS FOR THE ASSUMED CONSERVATIVENESS OF PI Better statistical methods for addressing missing cessation outcomes may not be widely adopted within the tobacco cessation research community because of the perceived conservativeness of this easily implemented PI approach. When rates of missing outcomes are the same among the different intervention groups, this approach does tend to be more conservative than the complete case analysis. However, this result is highly sensitive to the equivalence of the rates of missing outcomes. NONCONSERVATIVENESS OF PI IN APPLICATION PI is routinely criticized because of somewhat arbitrary performance and an overall lack of conservativeness in practice. Here, we present elementary statistical arguments demonstrating that, in commonly encountered situations, PI is as likely to lead to liberal estimates as to conservative estimates relative to the complete case analysis. DISCUSSION PI does not necessarily lead to more conservative or less biased effect estimates. Better statistical methods for addressing missing data need to be adopted within the tobacco cessation research community.

Annette Bar-Cohen

Papers

The impact of survey nonresponse bias on conclusions drawn from a mammography intervention trial