Richard M. Cullen

The criminal profiling illusion: what's behind the smoke and mirrors?

There is a belief that criminal profilers can predict a criminals characteristics from crime scene evidence. In this article, the authors argue that this belief may be an illusion and explain how people may have been misled into believing that criminal profiling (CP) works despite no sound theoretical grounding and no strong empirical support for this possibility. Potentially responsible for this illusory belief is the information that people acquire about CP, which is heavily influenced by anecdotes, repetition of the message that profiling works, the expert profiler label, and a disproportionate emphasis on correct predictions. Also potentially responsible are aspects of information processing such as reasoning errors, creating meaning out of ambiguous information, imitating good ideas, and inferring fact from fiction. The authors conclude that CP should not be used as an investigative tool because it lacks scientific support.

Implementation of the BETTER 2 program: a qualitative study exploring barriers and facilitators of a novel way to improve chronic disease prevention and screening in primary care.

BackgroundBETTER (Building on Existing Tools to Improve Chronic Disease Prevention and Screening in Primary Care) is a patient-based intervention to improve chronic disease prevention and screening (CDPS) for cardiovascular disease, diabetes, cancer, and associated lifestyle factors in patients aged 40 to 65. The key component of BETTER is a prevention practitioner (PP), a health care professional with specialized skills in CDPS who meets with patients to develop a personalized prevention prescription, using the BETTER toolkit and Brief Action Planning. The purpose of this qualitative study was to understand facilitators and barriers of the implementation of the BETTER 2 program among clinicians, patients, and stakeholders in three (urban, rural, and remote) primary care settings in Newfoundland and Labrador, Canada.MethodsWe collected and analyzed responses from 20 key informant interviews and 5 focus groups, as well as memos and field notes. Data were organized using Nvivo 10 software and coded using constant comparison methods. We then employed the Consolidated Framework for Implementation Research (CFIR) to focus our analysis on the domains most relevant for program implementation.ResultsThe following key elements, within the five CFIR domains, were identified as impacting the implementation of BETTER 2: (1) intervention characteristics—complexity and cost of the intervention; (2) outer setting—perception of fit including lack of remuneration, lack of resources, and duplication of services, as well as patients’ needs as perceived by physicians and patients; (3) characteristics of prevention practitioners—interest in prevention and ability to support and motivate patients; (4) inner setting—the availability of a local champion and working in a team versus working as a team; and (5) process—planning and engaging, collaboration, and teamwork.ConclusionsThe implementation of a novel CDPS program into new primary care settings is a complex, multi-level process. This study identified key elements that hindered or facilitated the implementation of the BETTER approach in three primary care settings in Newfoundland and Labrador. Employing the CFIR as an overarching typology allows for comparisons with other contexts and settings, and may be useful for practices, researchers, and policy-makers interested in the implementation of CDPS programs.

Hidden Markov model using Dirichlet process for de-identification

For the 2014 i2b2/UTHealth de-identification challenge, we introduced a new non-parametric Bayesian hidden Markov model using a Dirichlet process (HMM-DP). The model intends to reduce task-specific feature engineering and to generalize well to new data. In the challenge we developed a variational method to learn the model and an efficient approximation algorithm for prediction. To accommodate out-of-vocabulary words, we designed a number of feature functions to model such words. The results show the model is capable of understanding local context cues to make correct predictions without manual feature engineering and performs as accurately as state-of-the-art conditional random field models in a number of categories. To incorporate long-range and cross-document context cues, we developed a skip-chain conditional random field model to align the results produced by HMM-DP, which further improved the performance.

Outcomes of deliveries by family physicians or obstetricians: a population-based cohort study using an instrumental variable

Background: Previous research has suggested that obstetric outcomes are similar for deliveries by family physicians and obstetricians, but many of these studies were small, and none of them adjusted for unmeasured selection bias. We compared obstetric outcomes between these provider types using an econometric method designed to adjust for unobserved confounding. Methods: We performed a retrospective population-based cohort study of all Canadian (except Quebec) hospital births with delivery by family physicians and obstetricians at more than 20 weeks gestational age, with birth weight greater than 500 g, between Apr. 1, 2006, and Mar. 31, 2009. The primary outcomes were the relative risks of in-hospital perinatal death and a composite of maternal mortality and major morbidity assessed with multivariable logistic regression and instrumental variable–adjusted multivariable regression. Results: After exclusions, there were 3600 perinatal deaths and 14 394 cases of maternal morbidity among 799 823 infants and 793 053 mothers at 390 hospitals. For deliveries by family physicians v. obstetricians, the relative risk of perinatal mortality was 0.98 (95% confidence interval [CI] 0.85–1.14) and of maternal morbidity was 0.81 (95% CI 0.70–0.94) according to logistic regression. The respective relative risks were 0.97 (95% CI 0.58–1.64) and 1.13 (95% CI 0.65–1.95) according to instrumental variable methods. Interpretation: After adjusting for both observed and unobserved confounders, we found a similar risk of perinatal mortality and adverse maternal outcome for obstetric deliveries by family physicians and obstetricians. Whether there are differences between these groups for other outcomes remains to be seen.

Utility of computed tomography and derivation and validation of a score to identify an emergent outcome in 2,315 patients with suspected urinary tract stone.

OBJECTIVE Because a majority of urinary tract stones (UTSs) pass spontaneously and clinically significant alternative pathology is rare, we hypothesize that many computed tomographic (CT) scans to diagnose them are likely unnecessary. We sought to measure the impact of renal CT scans on resource use and to justify a prospective study to derive a score that predicts an emergent diagnosis in patients with suspected UTS by doing so in our retrospective series. METHODS We conducted a retrospective study of ED patients who had noncontrast CT of the abdomen for suspected UTS. A split-sample was used to derive and validate a score to predict the presence of an emergent diagnosis on CT. RESULTS Of the 2,315 patients (50.8% female, mean age 45 years), 49 (2.1%) had an emergent outcome observed on CT. An additional 12 (0.5%) patients had an urgent outcome and 239 (10.6%) had a urologic procedure within 8 weeks of the CT. Serum white blood cell count, highest temperature, urine red blood cell count, and the presence of abdominal pain were significant predictors of the primary outcome. A score derived using these predictors had a potential range of -2 (0.26% predicted risk, 0.5% actual risk of the outcome) to 6 (52% predicted risk). The score was moderately discriminatory with c-statistics of 0.752 (derivation) and 0.668 (validation) and accurate with Hosmer-Lemeshow statistics of 10.553 (p  =  0.228, derivation) and 9.70 (p  =  0.286, validation). CONCLUSIONS A sensible, relevant score derived and validated on all patients presenting with symptoms suggestive of renal colic could be useful in reducing abdominal CT scan ordering.

An Investigation of Cancer Rates in the Argentia Region, Newfoundland and Labrador: An Ecological Study

Background. The Argentia region of Newfoundland and Labrador, Canada, was home to a US naval base during a 40-year period between the 1940s and the 1990s. Activities on the base resulted in contamination of the soil and groundwater in the region with chemicals such as heavy metals and dioxins, and residents have expressed concern about higher rates of cancer in their community. This study investigated the rate of cancer diagnosis that is disproportionately high in the Argentia region. Methods. Cases of cancer diagnosed between 1985 and 2011 were obtained for the Argentia region, two comparison communities, and the province of Newfoundland and Labrador. Crude and age-standardized incidence rates of cancer diagnosis were calculated and compared. The crude incidence rate was adjusted for differences in age demographics using census data, and age-standardized incidence rates were compared. Results. Although the Argentia region had a higher crude rate of cancer diagnosis, the age-standardized incidence rate did not differ significantly from the comparison communities or the provincial average. Argentia has an aging population, which may have influenced the perception of increased cancer diagnosis in the community. Conclusions. We did not detect an increased burden of cancer in the Argentia region.