Sarcoidosis: a prospective observational cohort from Northern Alberta.

INTRODUCTION: Sarcoidosis is a multi-system disease reported to occur with a higher incidence in Alberta than many other health jurisdictions within and outside of Canada. The reasons for this higher incidence are currently not known. Exposure to beryllium can result in a clinically and radiologically identical disease to sarcoidosis. The purpose of our study was to identify patterns with potential occupational or environmental exposures to beryllium amongst individuals with sarcoidosis in Alberta through a tertiary referral center. METHODS: A prospective observational study was carried out at the University of Alberta Hospital. Patients with confirmed sarcoidosis (stages 0-4) were recruited from subspecialty clinics (Respirology, Cardiology, Neurology and Occupational Health). A predetermined list of industries thought to involve potentially relevant exposures for the development of sarcoidosis was used to capture current and previous exposure history. Results were entered into a database and where possible verified by comparing with existing electronic medical records (including histories, physical examination, diagnostic imaging and physiology). RESULTS: A total of 45 patients were recruited, 25 men and 20 women. Of these, 84% of participants reported working in or being exposed to an industry/environment suspected of contributing to development of sarcoidosis over their lifetime. The most frequently reported exposures were within farming and agriculture (27%), oil and gas (20%), metalworking and handling animals (18%). CONCLUSIONS: Amongst this cohort, a high proportion reported working with a potentially relevant exposure. Individuals being assessed for sarcoidosis should have their most responsible physician elicit a detailed work and environmental history. (Sarcoidosis Vasc Diffuse Lung Dis 2020; 37 (4): e2020014).

Routine use of post-bronchodilator testing in pulmonary function testing labs.

PURPOSE: Pulmonary function tests (PFTs), including spirometry with and without post-bronchodilator (post-BD) testing, are frequently performed in the assessment of asthma, along with other obstructive airway disorders. Multiple publications over the past 15 years have noted that one in three physician-diagnosed asthma cases are not in fact asthma. In this quality assurance project, we assess whether PFT labs in Alberta have policies on post-BD testing, as extraneous and unnecessary use of post-BD testing can lead to wasted staff and patient time and unnecessary expenses to the health care system. METHODS: We reviewed, in collaboration with the College of Physicians and Surgeons of Alberta and Alberta Medical Association, all PFT labs in the province of Alberta (hospital-based private not-for-profit [NFP] and private for-profit [FP] labs). This health policy study of PFT labs involved identifying the proportions and regional distribution of NFP and private FP labs in the province of Alberta while assessing post-BD policies. Each PFT lab was asked for their policy regarding spirometry and asthma diagnosis from May 1 to August 31, 2017. RESULTS: A total of 92 PFT labs were identified in Alberta, 74 of which were private FP (independent) labs, while 18 were private NFP (public) hospital-based labs. Policies were as follows: (i) post-BD policy existed (and if so routinely performed / not routinely done); (ii) no post-BD policy; and (iii) lab chose not to participate. All 18 hospital labs responded: 10 had no policy; six had a policy or algorithm; one did not perform post-BD testing (exercise testing) and one had multiple testing sites. Of the private FP labs, three had relevant policies and/or algorithm and 10 had none. No information was provided from 61 labs. Access to PFT labs in Northern Alberta was limited. CONCLUSIONS: Lab policies surrounding post-BD testing were found to be heterogeneous in Alberta. Low response rates, despite the use of a systems approach and requests in writing and in person from FP labs, were notable. Development of a standardized policy across the province would be beneficial. Further higher-level review of the appropriateness of post-BD use in both FP and NFP PFT labs is needed.

Psychiatric co-morbidity and asthma: A pilot study utilizing a free use tool to improve asthma care.

PURPOSE: To assess the prevalence of co-morbid psychiatric disorders in asthmatic patients in a Western Canadian Regional Severe Asthma Center. METHODS: A prospective study was completed of patients evaluated through the Edmonton Regional Severe Asthma Clinic (ERSAC). A standardised evaluation, the Mini International Neuropsychiatric Interview (MINI) screen was used to identify possible psychiatric disorders. RESULTS: Twenty-four individuals with moderate to severe asthma, who presented for treatment at ERSAC, were recruited and underwent assessment with the MINI screen. The average patient age was 48 years (range 18-81 years). Nine patients were male and fifteen were female. Twenty subjects (83%) screened positive for a possible psychiatric co-morbidity using the MINI screen. The most common psychiatric co-morbidities identified were post-traumatic stress disorder (50% of the sampled population), depressive episode or persistent depressive disorder (42%), substance/alcohol abuse (33%), generalized anxiety disorder (335), manic episode (25%), agoraphobia (21%), panic disorder (21%) and obsessive-compulsive disorder (17%). Some individuals had more than one concomitant possible psychiatric co-morbidity identified by the MINI screen. CONCLUSIONS: Psychiatric co-morbidity was confirmed to be common in patients with moderate-severe asthma. In individuals with asthma, the MINI screen appeared to be a simple and useful clinical tool to screen for untreated/sub-optimally-managed psychiatric co-morbidities that may impact management.

Neural network explanation using inversion.

An important drawback of many artificial neural networks (ANN) is their lack of explanation capability [Andrews, R., Diederich, J., & Tickle, A. B. (1996). A survey and critique of techniques for extracting rules from trained artificial neural networks. Knowledge-Based Systems, 8, 373-389]. This paper starts with a survey of algorithms which attempt to explain the ANN output. We then present HYPINV, a new explanation algorithm which relies on network inversion; i.e. calculating the ANN input which produces a desired output. HYPINV is a pedagogical algorithm, that extracts rules, in the form of hyperplanes. It is able to generate rules with arbitrarily desired fidelity, maintaining a fidelity-complexity tradeoff. To our knowledge, HYPINV is the only pedagogical rule extraction method, which extracts hyperplane rules from continuous or binary attribute neural networks. Different network inversion techniques, involving gradient descent as well as an evolutionary algorithm, are presented. An information theoretic treatment of rule extraction is presented. HYPINV is applied to example synthetic problems, to a real aerospace problem, and compared with similar algorithms using benchmark problems.