Area-based socio-economic disadvantage and tuberculosis incidence

OBJECTIVE

To examine neighborhood-level influences on tuberculosis (TB) incidence in a multilevel population-based sample.

DESIGN

All incident TB cases in Washington State, United States (n = 2161), reported between 1 January 2000 and 31 December 2008 were identified. Multivariate Poisson analysis was used at the ZIP Code tabulation area (ZCTA) level, which allowed for further exploration of area-specific influences on TB incidence.

RESULTS

A significant association was found between indices of socio-economic position (SEP) and TB incidence in Washington State, with a clear gradient of higher rates observed among lower ZCTA socio-economic quartiles. Compared to the wealthiest SEP quartile, the relative incidence of TB in successively lower quartiles was respectively 2.7, 4.1 and 10.4 (P trend <0.001). In multivariate analyses, the addition of area-level race, ethnicity and country of birth significantly attenuated this association (adjusted incidence rate ratios 2.3, 2.6, 5.7; P trend <0.001).

CONCLUSION

This study found a significant inverse association between area measures of socio-economic status (SES) and TB incidence across ZCTAs in Washington State, even after adjusting for individual age and sex and area-based race, ethnicity and foreign birth. These results emphasize the importance of neighborhood context and the need to target prevention efforts to low-SES neighborhoods.

Is household antibiotic use a risk factor for antibiotic-resistant pneumococcal infection?

We used microbiology and pharmacy data from health-maintenance organizations to determine whether antibiotic use by a household member increases the risk of penicillin-non-susceptible pneumococcal disease. Though it has been well established that an individual's antibiotic use increases one's risk of antibiotic-resistant infection, it is unclear whether the risk is increased if a member of one's household is exposed to antibiotics. We therefore conducted a case-control study of patients enrolled in health maintenance organizations in Western Washington and Northern California. Cases were defined as individuals with penicillin-non-susceptible pneumococcal infection; controls were individuals with penicillin-susceptible pneumococcal infection. Socioeconomic variables were obtained by linking addresses with 1997 census block group data. One-hundred and thirty-four cases were compared with 798 controls. Individual antibiotic use prior to diagnosis increased the odds of penicillin non-susceptibility, with the strongest effect seen for beta-lactam use within 2 months (OR 1.8, 95% CI 1.2, 2.8). When household antibiotic use by persons other than the patient were considered, at 4 months prior to diagnosis there was a trend towards an association between penicillin non-susceptibility and beta-lactam antibiotic use, and a possible association in a small subgroup of patients with eye and ear isolates. However, no significant overall pattern of association was seen. We conclude that though antibiotic use of any kind within 2 months prior to diagnosis is associated with an increased risk of penicillin-non-susceptible pneumococcal disease, there is no significant overall pattern of association between household antibiotic use and penicillin-non-susceptible pneumococcal infection.

Common bile duct injury during laparoscopic cholecystectomy and the use of intraoperative cholangiography: adverse outcome or preventable error?

BACKGROUND

Common bile duct (CBD) injury is a serious complication of laparoscopic cholecystectomy (LC). Predictors of this adverse outcome have not been well documented.

HYPOTHESIS

Surgeon experience and the use of intraoperative cholangiography (IOC) are associated with a decreased rate of major CBD injury during LC.

DESIGN

A retrospective population-based cohort study.

SETTING

Washington State hospital discharge database reports from 1991 through 1998.

PATIENTS

Discharge reports were searched for International Classification of Diseases, Ninth Revision, procedure codes consistent with LC and then evaluated for procedure codes for CBD repair and reconstruction within 90 days of LC.

MAIN OUTCOME MEASURE

The rate of CBD injury in patients undergoing LC based on the surgeon's experience and IOC use.

RESULTS

In all, 30 630 LCs and 76 major CBD injuries (2.5/1000 operations) were identified in this analysis. There were no significant differences between injured and noninjured patients in demographics, disease, payer status, or hospital variables. A CBD injury occurred in 3.2 of 1000 LCs in the early case order of surgeons compared with 1.7 per 1000 at later points (P = .01) (relative risk, 1.81; 95% confidence interval, 1.44-2.88). The rate of injury in LCs performed without IOC was 3.3 per 1000 compared with 2.0 per 1000 in LCs with IOC (P = .02) (relative risk, 1.7; 95% confidence interval, 1.1-2.6). Surgeon's experience and IOC use were independent predictors of injury.

CONCLUSIONS

The rate of CBD injury is significantly lower when IOC is used. This effect is magnified during the early experience of surgeons. Systematic use of IOC may significantly reduce the rate of CBD injury.

Has misdiagnosis of appendicitis decreased over time? A population-based analysis

CONTEXT

Misdiagnosis of presumed appendicitis is an adverse outcome that leads to unnecessary surgery. Computed tomography, ultrasonography, and laparoscopy have been suggested for use in patients with equivocal signs of appendicitis to decrease unnecessary surgery.

OBJECTIVE

To determine if frequency of misdiagnosis preceding appendectomy has decreased with increased availability of computed tomography, ultrasonography, and laparoscopy.

DESIGN, SETTING, AND PATIENTS

Retrospective, population-based cohort study of data from a Washington State hospital discharge database for 85 790 residents assigned International Classification of Diseases, Ninth Revision procedure codes for appendectomy, and United States Census Bureau data for 1987-1998.

MAIN OUTCOME MEASURE

Population-based age- and sex-standardized incidence of appendectomy with acute appendicitis (perforated or not) or with a normal appendix.

RESULTS

Among 63 707 nonincidental appendectomy patients, 84.5% had appendicitis (25.8% with perforation) and 15.5% had no associated diagnosis of appendicitis. After adjusting for age and sex, the population-based incidence of unnecessary appendectomy and of appendicitis with perforation did not change significantly over time. Among women of reproductive age, the population-based incidence of misdiagnosis increased 1% per year (P =.005). The incidence of misdiagnosis increased 8% yearly in patients older than 65 years (P<.001) but did not change significantly in children younger than 5 years (P =.17). The proportion of patients undergoing laparoscopic appendectomy who were misdiagnosed was significantly higher than that of open appendectomy patients (19.6% vs 15.5%; P<.001).

CONCLUSION

Contrary to expectation, the frequency of misdiagnosis leading to unnecessary appendectomy has not changed with the introduction of computed tomography, ultrasonography, and laparoscopy, nor has the frequency of perforation decreased. These data suggest that on a population level, diagnosis of appendicitis has not improved with the availability of advanced diagnostic testing.

Children hospitalized early and increased risk for future serious injury

OBJECTIVE

To determine if infants hospitalized for any reason before 90 days of age are at increased risk for future serious injury.

SETTING

Washington State.

METHODS

A population based retrospective cohort study, using data from Washington State birth and death certificates linked to a statewide hospital discharge database for the years 1989 through 1997. Participants included healthy full term infants born in Washington State between 1989 and 1995. A total of 29,466 infants hospitalized <90 days of age (early hospitalization) were compared to 29,750 randomly selected infants not hospitalized early. The primary outcome was an injury resulting in hospitalization or death between 3-24 months.

RESULTS

Among infants hospitalized early, 76/10,000 had a subsequent serious injury before age 2, compared with 47/ 10,000 infants without an early hospitalization (relative risk (RR) 1.6; 95% confidence interval (CI) 1.3 to 2.0). In a multivariate model including maternal age and parity, the adjusted RR for serious injury associated with early hospitalization was 1.5 (95% CI 1.2 to 1.8). Infants hospitalized early were three times as likely to be hospitalized between 3-24 months of age for intentional injury compared with infants not hospitalized early (RR 3.3; 95% CI 1.1 to 10.1).

CONCLUSIONS

Infants hospitalized in the first three months of life for any reason were 50% more likely to have a subsequent serious injury compared with infants not hospitalized early and were also at increased risk of intentional injury. This identifiable group of infants might be suitable for targeted childhood injury prevention programs including those involving prenatal and postnatal visits.

Anthropometric measures and risk of death in children with end-stage renal disease

We evaluated the association between anthropometric measurements and death among pediatric patients with end-stage renal disease (ESRD) using data from the Pediatric Growth and Development Special Study (PGDSS) from the US Renal Data System. Height, growth velocity, and body mass index (BMI) were used for the analysis of 1,949 patients in the PGDSS. To standardize these measurements, SD scores (SDSs) were calculated using population data from the Third National Health and Nutrition Examination Survey. Using Cox proportional hazards models, we assessed the association between anthropometric measures and death, controlling for demographic factors and stratifying by age. Multivariate analysis showed that each decrease by 1 SDS in height was associated with a 14% increase in risk for death (adjusted relative risk [aRR], 1.14; 95% confidence interval [CI], 1.02 to 1.27; P = 0.017). For each 1 SDS decrease in growth velocity among patients in our sample, the risk for death increased by 12% (aRR, 1.12; 95% CI, 1.00 to 1.25; P = 0.043). There was a statistically significant U-shaped association between BMI and death (P = 0.001), with relatively low and high BMIs associated with an increased risk for death. In children with ESRD, growth delay and extremes in BMI are associated with an increased risk for mortality.

The effect of recall on estimation of incidence rates for injury in Ghana

BACKGROUND

Injury is a major public health problem in many developing countries. Due to limitations of vital registry and health service data, surveys are an important tool to obtain information about injury in these countries. The value of such surveys can be limited by incomplete recall. The most appropriate recall period to use in surveys on injury in developing countries has not been well addressed.

METHODS

A household survey of injury in Ghana was conducted. Estimated annual non-fatal injury incidence rates were calculated for 12 recall periods (1-12 months prior to the interview, with each successively longer period including the preceding shorter periods).

RESULTS

There was a notable decline in the estimated rate from 27.6 per 100 per year for a one-month recall period to 7.6 per 100 per year for a 12-month recall period (72% decline). The extent of this decline was not influenced by age, gender, rural versus urban location, nor by type of respondent (in-person versus proxy). Rate of decline was influenced by severity of injury. Injuries resulting in <7 days of disability showed an 86% decline in estimated rates from a one-month to a 12-month recall period, whereas injuries resulting in > or =30 days of disability showed minimal decline.

CONCLUSIONS

In this setting, longer recall periods significantly underestimate the injury rate compared to shorter recall periods. Shorter recall periods (1-3 months) should be used when calculating the overall non-fatal injury incidence rate. However, longer recall periods (12 months) may be safely used to obtain information on the more severe, but less frequent, injuries.

Compliance with PASARR recommendations for Medicaid recipients in nursing homes. Preadmission Screening and Annual Resident Review

OBJECTIVE

To determine the rate of compliance with placement and mental health recommendations of the preadmission screening and annual resident review (PASARR) program.

DESIGN

A retrospective observational study using PASARR screening forms to identify recommendations of inpatient psychiatric care for people needing more care than nursing homes provide, alternate disposition for individuals needing less than nursing home level care, and recommendations for new mental health services when needed but not provided. Service use as indicated in Medicaid billing data during the following year was used to determine receipt of service.

PARTICIPANTS

All Washington state Medicaid recipients screened from 1992 through 1993.

MAIN OUTCOME MEASUREMENTS

Compliance rates for placement and service recommendations.

RESULTS

Inpatient psychiatric care was recommended for four of the 523 Medicaid recipients (0.8%), all of whom received it. Screeners recommended alternate dispositions in 131 (25%) subjects, and compliance occurred in 29% of these. Recommendations for new services were made in 310 (59%) cases. Compliance rates averaged 35%, ranging from 73% for medication recommendations to 7% for consultation. Depressed individuals were less likely to receive recommended services.

CONCLUSIONS

Many individuals needed additional mental health services but did not receive them, and a significant minority of patients could have been given an alternate disposition but rarely were.

Implementation of the Henry J. Kaiser Family Foundation's Community Health Promotion Grant Program: a process evaluation

The Community Health Promotion Grants Program, sponsored by the Henry J. Kaiser Family Foundation, represents a major health initiative that established 11 community health promotion projects. Successful implementation was characterized by several critical factors: (1) intervention activities; (2) community activation; (3) success in obtaining external funding; and (4) institutionalization. Analysis of the program was based on data from several sources: program reports, key informant surveys, and a community coalition survey. Results indicate that school-based programs focusing on adolescent health problems were the most successful in reaching the populations they were targeting. The majority of the programs were able to attract external funding, thereby adding to their initial resource base. The programs were less successful in generating health promotion activities and in achieving meaningful institutionalization in their communities.

Long-term survival of elderly trauma patients

OBJECTIVE

To evaluate the long-term survival and factors that influence survival among a cohort of elderly trauma patients compared with an uninjured cohort.

DESIGN

A retrospective cohort analysis.

DATA SOURCES

Health Care Finance Administration, Baltimore, Md, Medicare data.

SUBJECTS

A cohort of elderly patients (n = 9424) hospitalized for injury in 1987 was identified using Medicare hospital discharge abstract data. An uninjured comparison group (n = 37,787) was identified from Medicare eligibility files. For injured patients, an Injury Severity Score was generated from the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9CM) codes. For both cohorts, preexisting illness was assessed by ICD-9CM codes from Health Care Finance Administration outpatient and inpatient data files for 1986 and 1987.

MAIN OUTCOME MEASURES

Relative risk for mortality within 5 years subsequent to injury, adjusted for age, sex, and preexisting illness, using Cox proportional hazard regression.

RESULTS

The injured cohort had a significantly reduced 5-year survival when compared with the uninjured group (relative risk [RR] = 1.71; 95% confidence interval, 1.66-1.77). The lower survival persisted even among patients who survived at least 3 years after injury. Coexisting disease, age, and Injury Severity Score were strong predictors of survival.

CONCLUSIONS

The adverse effect of trauma on survival in elderly patients is not isolated to the immediate postinjury period, but lasts years after the trauma episode. Further study is required to identify the reasons for this persistent effect of trauma on subsequent survival.

Relationship between the volume of total hip replacements performed by providers and the rates of postoperative complications in the state of Washington

Since the late 1970's, an empirical relationship between the volume of procedures performed by a provider (a hospital or surgeon) and the outcome has been documented for various operations. The present study examines the relationship between the volume of hip replacements performed by surgeons and hospitals and the postoperative rate of complications. A statewide hospital discharge registry was used to identify patients who had had an elective hip replacement between 1988 and 1991. Patients who had had a revision procedure, who had been referred on an emergency basis, or who had had a diagnosis of a fracture or a malignant tumor on admission were excluded. There were 7936 eligible patients who had had 8774 hip replacements. The average annual number of all hip replacements performed from 1987 through 1991 was subsequently determined for each hospital and surgeon who had cared for at least one patient in the study cohort. The rate of operative complications was modeled as a function of the volume of procedures performed by the surgeon or hospital (the surgeon or hospital volume), with adjustment for the age of the patient, gender, co-morbidity, and operative diagnosis. We noted significant differences in the case mix of low-volume providers compared with that of high-volume providers (p < 0.01). In general, surgeons and hospitals with a volume below the fortieth percentile managed patients who had a more adverse risk profile in terms of age, co-morbidity, and diagnosis. Even after adjustment for the case mix, there was a significant relationship between surgeons who averaged fewer than two hip replacements annually (low-volume surgeons) and a worse outcome (p < 0.05). Patients managed by these low-volume surgeons tended to have higher mortality rates, more infections, higher rates of revision operations, and more serious complications during the index hospitalization. The duration of hospitalization was inversely related to surgeon volume and directly associated with hospital volume. Hospital charges were inversely related to hospital volume, even after adjustment for patient-related factors as well as the duration of hospitalization, the year of the operation, and the destination after discharge (p < 0.05). More detailed information is required to investigate the reason for these observed variations in the rates of complications. If future studies confirm an association between low-volume providers and an adverse outcome, performance of some types of elective total hip replacements at regional centers should be considered.

Trauma recidivism in the elderly

BACKGROUND

The incidence of recurrent trauma in the elderly is unknown. This study evaluated the risk of readmission for injury among elderly trauma patients compared with an uninjured geriatric cohort. The effects of age, sex, race, preexisting illness, and ISS on trauma recidivism were determined.

METHODS

Population based retrospective cohort analysis of the elderly using administrative data from the Health Care Financing Administration was performed. The measured outcome was trauma admission within 5 years. The injured were identified using hospital discharge data and the Injury Severity Score generated by ICD-Map. The uninjured were identified from Medicare eligibility files. Comorbid illness was assessed using ICD-9CM codes from outpatient and inpatient data files.

RESULTS

The injured members of the cohort had increased risk of subsequent new trauma admission (p < 0.001). Increasing Injury Severity Score, age, and comorbid illness are associated with trauma recidivism.

CONCLUSION

Trauma in the elderly is recurrent. Further study is required to develop age and injury specific interventions to prevent recurrent injury.

Hospitalizations for injury in New Zealand: prior injury as a risk factor for assaultive injury

OBJECTIVES

This study sought to determine the degree to which injury hospitalization, especially for assaultive injury, is a risk for subsequent hospitalization due to assault.

METHODS

A New Zealand hospitalization database was used to perform a retrospective cohort study. Exposure was defined as an injury hospitalization, stratified into assaultive and nonassaultive mechanisms. Hospitalizations for an assault during a 12-month follow-up period were measured.

RESULTS

Individuals with a prior nonassaultive injury were 3.2 times more likely to be admitted for an assault than those with no injury admission (95% confidence interval [CI] = 2.7, 3.9). The relative risk associated with a prior assault was 39.5 (95% CI = 35.8, 43.5), and the subsequent admission rate did not vary significantly by sex, race, or marital or employment status. Among those readmitted for an assault, 70% were readmitted within 30 days of the initial hospitalization.

CONCLUSIONS

Prior injury is a risk for serious assault, and the risk is even greater if the injury is due to assault. Risk of readmission for assault is largely independent of demographic factors and greatest within 30 days of the initial assault.

Motorcycle helmet use and injury outcome and hospitalization costs from crashes in Washington State

OBJECTIVES

The incidence, type, severity, and costs of crash-related injuries requiring hospitalization or resulting in death were compared for helmeted and unhelmeted motorcyclists.

METHODS

This was a retrospective cohort study of injured motorcyclists in Washington State in 1989. Motorcycle crash data were linked to statewide hospitalization and death data.

RESULTS

The 2090 crashes included in this study resulted in 409 hospitalizations (20%) and 59 fatalities (28%). Although unhelmeted motorcyclists were only slightly more likely to be hospitalized overall, they were more severely injured, nearly three times more likely to have been head injured, and nearly four times more likely to have been severely or critically head injured than helmeted riders. Unhelmeted riders were also more likely to be readmitted to a hospital for follow-up treatment and to die from their injuries. The average hospital stay for unhelmeted motorcyclists was longer, and cost more per case; the cost of hospitalization for unhelmeted motorcyclists was 60% more overall ($3.5 vs $2.2 million).

CONCLUSIONS

Helmet use is strongly associated with reduced probability and severity of injury, reduced economic impact, and a reduction in motorcyclist deaths.

Optimal survey design for community intervention evaluations: cohort or cross-sectional?

Community intervention evaluations that measure changes over time may conduct repeated cross-sectional surveys, follow a cohort of residents over time, or (often) use both designs. Each survey design has implications for precision and cost. To explore these issues, we assume that two waves of surveys are conducted, and that the goal is to estimate change in behavior for people who reside in the community at both times. Cohort designs are shown to provide more accurate estimates (in the sense of lower mean squared error) than cross-sectional estimates if (1) there is strong correlation over time in an individual's behavior at time 0 and time 1, (2) relatively few subjects are lost to followup, (3) the bias is relatively small, and (4) the available sample size is not too large. Otherwise, a repeated cross-sectional design is more efficient. We developed methods for choosing between the two designs, and applied them to actual survey data. Owing to drop-outs and losses to followup, the cohort estimates were usually more biased than the cross-sectional estimates. The correlations over time for most of the variables studied were also high. In many instances the cohort estimate, although biased, is preferred to the relatively unbiased cross-sectional estimate because the mean squared error was smaller for the cohort than for the cross-sectional estimate. If these results are replicated in other data, they may result in guidelines for choosing a more efficient study design.

Breaking the matches in a paired t-test for community interventions when the number of pairs is small

There is considerable interest in community interventions for health promotion, where the community is the experimental unit. Because such interventions are expensive, the number of experimental units (communities) is usually small. Because of the small number of communities involved, investigators often match treatment and control communities on demographic variables before randomization to minimize the possibility of a bad split. Unfortunately, matching has been shown to decrease the power of the design when the number of pairs is small, unless the matching variable is very highly correlated with the outcome variable (in this case, with change in the health behaviour). We used computer simulation to examine the performance of an approach in which we matched communities but performed an unmatched analysis. If the appropriate matching variables are unknown, and there are fewer than ten pairs, an unmatched design and analysis has the most power. If, however, one prefers a matched design, then for N < 10, power can be increased by performing an unmatched analysis of the matched data. We also discuss a variant of this procedure, in which an unmatched analysis is performed only if the matching 'did not work'.

Buffered lidocaine: analgesia for intravenous line placement in children

OBJECTIVES

To evaluate the effectiveness of intradermal buffered lidocaine as analgesia before intravenous line (i.v.) placement in children.

METHODS

This was a randomized clinical trial undertaken in the emergency department (ED) of a regional children's hospital. Participants were children 8 to 15 years old, seen in the ED and in need of i.v. lines. They were enrolled by three ED nurses. Participants were randomized to receive either intradermal buffered lidocaine or no analgesia. Before placement of the i.v. line, patients recorded the amount of pain they were in (baseline pain) on a visual analog pain scale. The primary outcome measure was amount of pain caused by the initial i.v. attempt, even if that attempt was unsuccessful. This was recorded by the participant on a visual analog scale. Demographic characteristics, the number of attempts to successful placement, and the time required to place the i.v. line were also recorded. Differences in pain of initial i.v. attempt and time to place the i.v. line were evaluated with the Mann-Whitney U test. Differences in success of i.v. line placement were evaluated with the chi 2 test.

RESULTS

Fifty-nine patients completed the study. Thirty received buffered lidocaine, and 29 received no analgesia before i.v. line placement. There was no significant difference between the two groups with regard to baseline pain or demographic characteristics. The median level of pain of the initial i.v. attempt as measured by the visual analog scale was 2.3 in the buffered-lidocaine group and 4.4 in the no-lidocaine group. Thirty-three percent of patients in the lidocaine group and 28% percent in the no-lidocaine group required more than one i.v. attempt. The median time to i.v. line placement was 10 minutes in the lidocaine group and 6 minutes in the no-lidocaine group.

CONCLUSIONS

Use of intradermal buffered lidocaine is an effective way to diminish the pain of i.v. line placement in children 8 to 15 years of age. There was no difference in i.v. success rate in this study; however, larger numbers of patients would be required to detect statistically significant differences. We recommend the routine use of intradermal buffered lidocaine for analgesia before i.v. line placement in older children in all but emergent situations.

Evaluating community-based nutrition programs: comparing grocery store and individual-level survey measures of program impact

BACKGROUND

This paper examines whether an "environmental indicator"--a survey of grocery store product displays--can provide a realistic alternative to individual-level telephone surveys for the evaluation of community-based nutrition programs.

METHODS

Telephone and grocery store measures were used separately to evaluate three community-level dietary interventions that were part of the Kaiser Family Foundation Community Health Promotion Grants Program (CHPGP). Both surveys were conducted in the three intervention and seven control communities at three points in time: 1988, 1990, and 1992. The grocery store survey recorded the relative availability of low-fat and high-fiber products and the amount of store-provided health-education information. Self-reported dietary intake of residents was obtained in the same communities using a telephone survey.

RESULTS

In the one community in which the intervention seemed to have contributed to reduced fat consumption, the grocery store and telephone surveys showed very similar relative changes for the only variable they had in common, low-fat milk consumption. In another community, both survey approaches indicated that there was no change or perhaps a slight worsening in the treatment relative to the controls. The third community produced the only contradictory results: the telephone survey suggested no change or perhaps a worsening, while the grocery store results were generally positive, though not statistically significant.

CONCLUSION

These results, combined with the much lower cost of the grocery store survey, justify further pursuit of environmental indicators as an evaluation tool.

A community-based approach to preventing alcohol use among adolescents on an American Indian reservation

This paper examines the effectiveness of a 5-year community-based health promotion program to reduce the rate of substance use, particularly alcohol, by adolescents on a Plains State American Indian reservation. The program was part of the Kaiser Family Foundation Community Health Promotion Grants Program. Since a reservation control group was not available, adolescents serving as control groups for other Community Health Promotion Grants Program communities, including a small sample of rural American Indians, were used as a basis for comparison. School-based surveys of 9th and 12th graders were carried out on the reservation and in five relevant California control communities--two suburban, three rural--in 1988, 1990, and 1992. The results showed that the use of both alcohol and marijuana declined substantially among American Indian adolescents living on the reservation. Binge drinking (five or more drinks on an occasion) declined from 46 percent to 30 percent, and marijuana use (in the past month) declined from 46 percent to 29 percent over the 4-year period. However, there were similar, if smaller, declines in alcohol use in the comparison groups. Since there was no evidence of a relative increase in exposure to alcohol and drug programs on the reservation, the authors are cautious in attributing the significant and heartening declines in substance use among adolescents on the reservation to the community-based program.

Assessing the validity of a survey of the restaurant health promotion environment

Our primary question was whether a telephone survey of restaurant personnel could provide accurate community-level measures of the restaurant health promotion environment. An obvious concern was that restaurant personnel might exaggerate the extent to which their establishment had a positive health promotion environment. Comparison with the most obvious "gold standard"--direct observation--showed fairly accurate reporting about nonsmoking seating arrangements, but restaurant personnel exaggerated the extent to which menu items were designated as low in fat. We also compared the restaurant-survey measures of nonsmoking seating availability at the community level with measures of the restrictiveness of local no-smoking ordinances. We found a positive relationship, as expected, between measures of the restrictiveness of ordinances and the amount of nonsmoking seating indicated by the restaurant survey.