Potentially ineffective care: time for earnest reexamination

Invasive Mechanical Ventilation in California Over 2000-2009: Implications for Emergency Medicine

INTRODUCTION

Patients who require invasive mechanical ventilation (IMV) often represent a sequence of care between the emergency department (ED) and intensive care unit (ICU). Despite being the most populous state, little information exists to define patterns of IMV use within the state of California.

METHODS

We examined data from the masked Patient Discharge Database of California's Office of Statewide Health Planning and Development from 2000-2009. Adult patients who received IMV during their stay were identified using the International Classification of Diseases 9th Revision and Clinical Modification procedure codes (96.70, 96.71, 96.72). Patients were divided into age strata (18-34 yr, 35-64 yr, and >65 yr). Using descriptive statistics and regression analyses, for IMV discharges during the study period, we quantified the number of ED vs. non-ED based admissions; changes in patient characteristics and clinical outcome; evaluated the marginal costs for IMV; determined predictors for prolonged acute mechanical ventilation (PAMV, i.e. IMV>96 hr); and projected the number of IMV discharges and ED-based admissions by year 2020.

RESULTS

There were 696,634 IMV discharges available for analysis. From 2000-2009, IMV discharges increased by 2.8%/year: n=60,933 (293/100,000 persons) in 2000 to n=79,868 (328/100,000 persons) in 2009. While ED-based admissions grew by 3.8%/year, non-ED-based admissions remained stable (0%). During 2000-2009, fastest growth was noted for 1) the 35-64 year age strata; 2) Hispanics; 3) patients with non-Medicare public insurance; and 4) patients requiring PAMV. Average total patient cost-adjusted charges per hospital discharge increased by 29% from 2000 (from $42,528 to $60,215 in 2014 dollars) along with increases in the number of patients discharged to home and skilled nursing facilities. Higher marginal costs were noted for younger patients (ages 18-34 yr), non-whites, and publicly insured patients. Some of the strongest predictors for PAMV were age 35-64 years (OR=1.12; 95% CI [1.09-1.14], p<0.05); non-Whites; and non-Medicare public insurance. Our models suggest that by 2020, IMV discharges will grow to n=153,153 (377 IMV discharges/100,000 persons) with 99,095 admitted through the ED.

CONCLUSION

Based on sustained growth over the past decade, by the year 2020, we project a further increase to 153,153 IMV discharges with 99,095 admitted through the ED. Given limited ICU bed capacities, ongoing increases in the number and type of IMV patients have the potential to adversely affect California EDs that often admit patients to ICUs.

Using conjoint analysis to develop a system of scoring policymakers' use of research in policy and program development

Background

The importance of utilising the best available research evidence in the development of health policies, services, and programs is increasingly recognised, yet few standardised systems for quantifying policymakers’ research use are available. We developed a comprehensive measurement and scoring tool that assesses four domains of research use (i.e. instrumental, conceptual, tactical, and imposed). The scoring tool breaks down each domain into its key subactions like a checklist. Our aim was to develop a tool that assigned appropriate scores to each subaction based on its relative importance to undertaking evidence-informed health policymaking. In order to establish the relative importance of each research use subaction and generate this scoring system, we conducted conjoint analysis with a sample of knowledge translation experts.

Methods

Fifty-four experts were recruited to undertake four choice surveys. Respondents were shown combinations of research use subactions called profiles, and rated on a 1 to 9 scale whether each profile represented a limited (1–3), moderate (4–6), or extensive (7–9) example of research use. Generalised Estimating Equations were used to analyse respondents’ choice data, which calculated a utility coefficient for each subaction. A large utility coefficient indicated that a subaction was particularly influential in guiding experts’ ratings of extensive research use.

Results

Utility coefficients were calculated for each subaction, which became the points assigned to the subactions in the scoring system. The following subactions yielded the largest utilities and were regarded as the most important components of each research use domain: using research to directly influence the core of the policy decision; using research to inform alternative perspectives to deal with the policy issue; using research to persuade targeted stakeholders to support a predetermined decision; and using research because it was a mandated requirement by the policymaker’s organisation.

Conclusions

We have generated an empirically derived and context-sensitive means of measuring and scoring the extent to which policymakers used research to inform the development of a policy document. The scoring system can be used by organisations to not only quantify the extent of their research use, but also to provide them with insights into potential strategies to improve subsequent research use.

Electronic supplementary material

The online version of this article (doi:10.1186/s12961-015-0022-y) contains supplementary material, which is available to authorized users.