Standard inference procedures for regression analysis make assumptions that are rarely satisfied in practice. Adjustments must be made to insure the validity of statistical inference. These adjustments, known for many years, are used routinely by some health researchers but not by others. We review some of these methods and give an example of their use in a health services study for a continuous and a count outcome. For the continuous outcome, we describe retransformation using the smear factor, accounting for missing cases via multiple imputation and attrition weights and improving results with bootstrap methods. For the count outcome, we describe zero inflated Poisson and negative binomial models and the two-part model to account for overabundance of zero values. Recent advances in computing and software development have produced user-friendly computer programs that enable the data analyst to improve prediction and inference based on regression analysis.

Background Animal-assisted therapy improves physiological and psychosocial variables in healthy and hypertensive patients. Objectives To determine whether a 12-minute hospital visit with a therapy dog improves hemodynamic measures, lowers neurohormone levels, and decreases state anxiety in patients with advanced heart failure. Methods A 3-group randomized repeated-measures experimental design was used in 76 adults. Longitudinal analysis was used to model differences among the 3 groups at 3 times. One group received a 12-minute visit from a volunteer with a therapy dog; another group, a 12-minute visit from a volunteer; and the control group, usual care. Data were collected at baseline, at 8 minutes, and at 16 minutes. Results Compared with controls, the volunteer-dog group had significantly greater decreases in systolic pulmonary artery pressure during (−4.32 mm Hg, P = .03) and after (−5.78 mm Hg, P = .001) and in pulmonary capillary wedge pressure during (−2.74 mm Hg, P = .01) and after (−4.31 mm Hg, P = .001) the intervention. Compared with the volunteer-only group, the volunteer-dog group had significantly greater decreases in epinephrine levels during (−15.86 pg/mL, P = .04) and after (−17.54 pg/mL, P = .04) and in norepinephrine levels during (−232.36 pg/mL, P = .02) and after (−240.14 pg/mL, P = .02) the intervention. After the intervention, the volunteer-dog group had the greatest decrease from baseline in state anxiety sum score compared with the volunteer-only (−6.65 units, P =.002) and the control groups (−9.13 units, P < .001). Conclusions Animal-assisted therapy improves cardiopulmonary pressures, neurohormone levels, and anxiety in patients hospitalized with heart failure.

Aims and objectives. To describe the frequency of nurses' delivery of tobacco cessation interventions ('Five A's': Ask, Advise, Assess, Assist, Arrange) and to determine the relationship of interventions to nurses' awareness of the Tobacco Free Nurses initiative. Background. Tobacco cessation interventions can be effectively provided by nurses. The delivery of smoking cessation interventions by healthcare providers is mandated by several organisations in the USA and around the world. Lack of education and resources about tobacco cessation may contribute to the minimal level of interventions. The Tobacco Free Nurses initiative was developed to provide nurses with easy access to web-based resources about tobacco control. Design. Cross-sectional survey of nurses (n = 3482) working in 35 Magnet-designated hospitals in the USA (21% response rate). Method. A valid and reliable questionnaire used in previous studies to assess the frequency of the nurse's delivery of smoking cessation interventions ('Five A's') was adapted for use on the web. Results. The majority of nurses asked (73%) and assisted (73%) with cessation. However, only 24% recommended pharmacotherapy. Only 22% referred to community resources and only 10% recommended use of the quitline. Nurses familiar with TFN (15%) were significantly more likely to report delivery of all aspects of interventions, including assisting with cessation (OR = 1AE55, 95% CI 1AE27, 1AE90) and recommending medications (OR = 1AE81, 95% CI 1AE45, 2AE24). Conclusions. Nurses' delivery of comprehensive smoking cessation interventions was suboptimal. Awareness of Tobacco Free Nurses was associated with increased interventions. Relevance to clinical practice. Further efforts are needed to ensure that nurses incorporate evidence-based interventions into clinical practice to help smokers quit. These findings support the value of Tobacco Free Nurses in providing nurses with information to support patients' quit attempts.