Medical Diagnosis Reimagined as a Process of Bayesian Reasoning and Elimination

This article delves into the interface between the art of medical diagnosis and the mathematical foundations of probability, the Bayes theorem. In a healthcare ecosystem witnessing an artificial intelligence (AI)-driven transformation, understanding the convergence becomes crucial for physicians. Contrary to viewing AI as a mysterious "black box," we demonstrate how every diagnostic decision by a medical practitioner is, in essence, Bayesian reasoning in action. The Bayes theorem is a mathematical translation of systematically updating our belief: it quantifies how an additional piece of information updates our prior belief in something. Using a clinical scenario of Kartagener syndrome, we showcase the parallels between a physician's evolving diagnostic thought process and the mathematical updating of prior beliefs with new evidence. By reimagining medical diagnosis through the lens of Bayes, this paper aims to demystify AI, accentuating its potential role as an enhancer of clinical acumen rather than a replacement. The ultimate vision presented is one of harmony, where AI serves as a symbiotic partner to physicians, with the shared goal of holistic patient care.

Quality improvement strategies for diabetes care: Effects on outcomes for adults living with diabetes

BACKGROUND

There is a large body of evidence evaluating quality improvement (QI) programmes to improve care for adults living with diabetes. These programmes are often comprised of multiple QI strategies, which may be implemented in various combinations. Decision-makers planning to implement or evaluate a new QI programme, or both, need reliable evidence on the relative effectiveness of different QI strategies (individually and in combination) for different patient populations.

OBJECTIVES

To update existing systematic reviews of diabetes QI programmes and apply novel meta-analytical techniques to estimate the effectiveness of QI strategies (individually and in combination) on diabetes quality of care.

SEARCH METHODS

We searched databases (CENTRAL, MEDLINE, Embase and CINAHL) and trials registers (ClinicalTrials.gov and WHO ICTRP) to 4 June 2019. We conducted a top-up search to 23 September 2021; we screened these search results and 42 studies meeting our eligibility criteria are available in the awaiting classification section.

SELECTION CRITERIA

We included randomised trials that assessed a QI programme to improve care in outpatient settings for people living with diabetes. QI programmes needed to evaluate at least one system- or provider-targeted QI strategy alone or in combination with a patient-targeted strategy. - System-targeted: case management (CM); team changes (TC); electronic patient registry (EPR); facilitated relay of clinical information (FR); continuous quality improvement (CQI). - Provider-targeted: audit and feedback (AF); clinician education (CE); clinician reminders (CR); financial incentives (FI). - Patient-targeted: patient education (PE); promotion of self-management (PSM); patient reminders (PR). Patient-targeted QI strategies needed to occur with a minimum of one provider or system-targeted strategy.

DATA COLLECTION AND ANALYSIS

We dual-screened search results and abstracted data on study design, study population and QI strategies. We assessed the impact of the programmes on 13 measures of diabetes care, including: glycaemic control (e.g. mean glycated haemoglobin (HbA1c)); cardiovascular risk factor management (e.g. mean systolic blood pressure (SBP), low-density lipoprotein cholesterol (LDL-C), proportion of people living with diabetes that quit smoking or receiving cardiovascular medications); and screening/prevention of microvascular complications (e.g. proportion of patients receiving retinopathy or foot screening); and harms (e.g. proportion of patients experiencing adverse hypoglycaemia or hyperglycaemia). We modelled the association of each QI strategy with outcomes using a series of hierarchical multivariable meta-regression models in a Bayesian framework. The previous version of this review identified that different strategies were more or less effective depending on baseline levels of outcomes. To explore this further, we extended the main additive model for continuous outcomes (HbA1c, SBP and LDL-C) to include an interaction term between each strategy and average baseline risk for each study (baseline thresholds were based on a data-driven approach; we used the median of all baseline values reported in the trials). Based on model diagnostics, the baseline interaction models for HbA1c, SBP and LDL-C performed better than the main model and are therefore presented as the primary analyses for these outcomes. Based on the model results, we qualitatively ordered each QI strategy within three tiers (Top, Middle, Bottom) based on its magnitude of effect relative to the other QI strategies, where 'Top' indicates that the QI strategy was likely one of the most effective strategies for that specific outcome. Secondary analyses explored the sensitivity of results to choices in model specification and priors.  Additional information about the methods and results of the review are available as Appendices in an online repository. This review will be maintained as a living systematic review; we will update our syntheses as more data become available.

MAIN RESULTS

We identified 553 trials (428 patient-randomised and 125 cluster-randomised trials), including a total of 412,161 participants. Of the included studies, 66% involved people living with type 2 diabetes only. Participants were 50% female and the median age of participants was 58.4 years. The mean duration of follow-up was 12.5 months. HbA1c was the commonest reported outcome; screening outcomes and outcomes related to cardiovascular medications, smoking and harms were reported infrequently. The most frequently evaluated QI strategies across all study arms were PE, PSM and CM, while the least frequently evaluated QI strategies included AF, FI and CQI. Our confidence in the evidence is limited due to a lack of information on how studies were conducted.  Four QI strategies (CM, TC, PE, PSM) were consistently identified as 'Top' across the majority of outcomes. All QI strategies were ranked as 'Top' for at least one key outcome. The majority of effects of individual QI strategies were modest, but when used in combination could result in meaningful population-level improvements across the majority of outcomes. The median number of QI strategies in multicomponent QI programmes was three.  Combinations of the three most effective QI strategies were estimated to lead to the below effects:  - PR + PSM + CE: decrease in HbA1c by 0.41% (credibility interval (CrI) -0.61 to -0.22) when baseline HbA1c < 8.3%; - CM + PE + EPR: decrease in HbA1c by 0.62% (CrI -0.84 to -0.39) when baseline HbA1c > 8.3%;  - PE + TC + PSM: reduction in SBP by 2.14 mmHg (CrI -3.80 to -0.52) when baseline SBP < 136 mmHg; - CM + TC + PSM: reduction in SBP by 4.39 mmHg (CrI -6.20 to -2.56) when baseline SBP > 136 mmHg;  - TC + PE + CM: LDL-C lowering of 5.73 mg/dL (CrI -7.93 to -3.61) when baseline LDL < 107 mg/dL; - TC + CM + CR: LDL-C lowering by 5.52 mg/dL (CrI -9.24 to -1.89) when baseline LDL > 107 mg/dL. Assuming a baseline screening rate of 50%, the three most effective QI strategies were estimated to lead to an absolute improvement of 33% in retinopathy screening (PE + PR + TC) and 38% absolute increase in foot screening (PE + TC + Other).

AUTHORS' CONCLUSIONS

There is a significant body of evidence about QI programmes to improve the management of diabetes. Multicomponent QI programmes for diabetes care (comprised of effective QI strategies) may achieve meaningful population-level improvements across the majority of outcomes. For health system decision-makers, the evidence summarised in this review can be used to identify strategies to include in QI programmes. For researchers, this synthesis identifies higher-priority QI strategies to examine in further research regarding how to optimise their evaluation and effects. We will maintain this as a living systematic review.

[Performance evaluation of a precision dosing service for vancomycin in a tertiary level pediatric hospital]

Background

Plasma level-based therapeutic drug monitoring of vancomycin is recommended in the treatment of complex pediatric infections in order to increase the probability of achieving safe and effective pharmacotherapy.

Objective

To retrospectively evaluate the activities and performance of pharmacotherapeutic optimization based on vancomycin levels at a tertiary pediatric hospital between 2007 and 2020.

Métodos

Vancomycin levels of pediatric patients were analyzed, assessing care quality indicators and analytical verifications, as well as aspects related to teaching and research. The predictive performance of vancomycin levels was evaluated after adjustment of the therapeutic regimen using a population pharmacokinetic optimization program (BestDose v1.126) considering the coefficient of determination (R2), the mean absolute percentage error (MAPE), and the root mean square error (RMSE).

Results

13269 vancomycin level determinations were analyzed; 70% were trough levels and 81% belonged to patients in the intensive care units. Forty percent of the trough levels were within the therapeutic range when adjusted without software. Three hundred seventy-four pharmacotherapeutic interventions, of which 97% were accepted by the treating physician; 75% of the post-adjustment trough levels were within the therapeutic range, compared to 40% when the approach was empirical, a difference that was statistically significant (p=0.03). The values associated with predictive performance (n subgroup of patients = 91) were: R2=0.61, MAPE=28.16%, and RMSE=3.3, which all showed to be adequate.

Conclusion

The performance of therapeutic vancomycin monitoring and related pharmacokinetic clinical activities showed to be good.

Validation of RT-qPCR test for SARS-CoV-2 in saliva specimens

Background

Saliva samples may be an easier, faster, safer, and cost-saving alternative to NPS samples, and can be self-collected by the patient. Whether SARS-CoV-2 RT-qPCR in saliva is more accurate than in nasopharyngeal swaps (NPS) is uncertain. We evaluated the accuracy of the RT-qPCR in both types of samples, assuming both approaches were imperfect.

Methods

We assessed the limit of detection (LoD) of RT-qPCR in each type of sample. We collected paired NPS and saliva samples and tested them using the Berlin Protocol to detect SARS-CoV-2 envelope protein (E). We used a Bayesian latent class analysis (BLCA) to estimate the sensitivity and specificity of each test, while accounting for their conditional dependence.

Results

The LoD were 10 copies/mL in saliva and 100 copies/mL in NPS. Paired samples of saliva and NPS were collected in 412 participants. Out of 68 infected cases, 14 were positive only in saliva. RT-qPCR sensitivity ranged from 82.7% (95% CrI: 54.8, 94.8) in NPS to 84.5% (50.9, 96.5) in saliva. Corresponding specificities were 99.1 (95% CrI: 95.3, 99.8) and 98.4 (95% CrI: 92.8, 99.7).

Conclusions

SARS-CoV-2 RT-qPCR test in saliva specimens has a similar or better accuracy than RT-qPCR test in NPS. Saliva specimens may be ideal for surveillance in general population, particularly in children, and in healthcare or other personnel in need of serial testing.

Radiography and Clinical Decision-Making in Chiropractic

The concern over x-ray exposure risks can overshadow the potential benefit of radiography, especially in cases where manual therapy is employed. Spinal malalignment cannot be accurately visualized without imaging. Manual therapy and the load tolerances of injured spinal tissues raise different criteria for the use of x-rays for spinal disorders than in medical practice. Current regulatory bodies rely on radiography risk assessments based on Linear-No-Threshold (LNT) risk models. There is a need to consider radiography guidelines for chiropractic which are different from those for medical practice. Radiography practice guidelines are summaries dominated by frequentist interpretations in the analysis of data from studies. In contrast, clinicians often employ a pseudo-Bayesian form of reasoning during the clinical decision-making process. The overrepresentation of frequentist perspectives in evidence-based practice guidelines alter decision-making away from practical assessment of a patient's needs, toward an overly cautious standard applied to patients without regard to their risk/benefit likelihoods relating to radiography. Guidelines for radiography in chiropractic to fully assess the condition of the spine and spinal alignment prior to manual therapy, especially with high velocity, low amplitude spinal manipulation (HVLA-SM), should necessarily differ from those used in medical practice.

The computation of case fatality rate for novel coronavirus (COVID-19) based on Bayes theorem: An observational study

BACKGROUND

When a new disease such starts to spread, the commonly asked questions are how deadly is it? and how many people are likely to die of this outbreak? The World Health Organization (WHO) announced in a press conference on January 29, 2020 that the death rate of COVID-19 was 2% on the case fatality rate (CFR). It was underestimated assuming no lag days from symptom onset to deaths while many CFR formulas have been proposed, the estimation on Bays theorem is worthy of interpretation. Hence, it is hypothesized that the over-loaded burdens of treating patients and capacities to contain the outbreak (LSBHRS) may increase the CFR.

METHODS

We downloaded COVID-19 outbreak numbers from January 21 to February 14, 2020, in countries/regions on a daily basis from Github that contains information on confirmed cases in >30 Chinese locations and other countries/regions. The pros and cons were compared among the 5 formula of CFR, including [A] deaths/confirmed; [B] deaths/(deaths + recovered); [C] deaths/(cases x days ago); [D] Bayes estimation based on [A] and the outbreak (LSBHRS) in each country/region; and [E] Bayes estimation based on [C] deaths/(cases x days ago). The coefficients of variance (CV = the ratio of the standard deviation to the mean) were applied to measure the relative variability for each CFR. A dashboard was developed for daily display of the CFR across each region.

RESULTS

The Bayes based on (A)[D] has the lowest CV (=0.10) followed by the deaths/confirmed (=0.11) [A], deaths/(deaths + recoveries) (=0.42) [B], Bayes based on (C) (=0.49) [E], and deaths/(cases x days ago) (=0.59) [C]. All final CFRs will be equal using the formula (from, A to E). A dashboard was developed for the daily reporting of the CFR. The CFR (3.7%) greater than the prior CFR of 2.2% was evident in LSBHRS, increasing the CFR. A dashboard was created to present the CFRs on COVID-19.

CONCLUSION

We suggest examining both trends of the Bayes based on both deaths/(cases 7 days ago) and deaths/confirmed cases as a reference to the final CFR. An app developed for displaying the provisional CFR with the 2 CFR trends can improve the underestimated CFR reported by WHO and media.

A Bayesian Approach to Understanding Sex Differences in Tuberculosis Disease Burden

Globally, men have a higher epidemiologic burden of tuberculosis (incidence, prevalence, mortality) than women do, possibly due to differences in disease incidence, treatment initiation, self-cure, and/or untreated-tuberculosis mortality rates. Using a simple, sex-stratified compartmental model, we employed a Bayesian approach to explore which factors most likely explain men's higher burden. We applied the model to smear-positive pulmonary tuberculosis in Vietnam (2006-2007) and Malawi (2013-2014). Posterior estimates were consistent with sex-specific prevalence and notifications in both countries. Results supported higher incidence in men and showed that both sexes faced longer durations of untreated disease than estimated by self-reports. Prior untreated disease durations were revised upward 8- to 24-fold, to 2.2 (95% credible interval: 1.7, 2.9) years for men in Vietnam and 2.8 (1.8, 4.1) years for men in Malawi, approximately a year longer than for women in each country. Results imply that substantial sex differences in tuberculosis burden are almost solely attributable to men's disadvantages in disease incidence and untreated disease duration. The latter, for which self-reports provide a poor proxy, implies inadequate coverage of case-finding strategies. These results highlight an urgent need for better understanding of gender-related barriers faced by men and support the systematic targeting of men for screening.

Final diagnoses and probability of new reason-for-encounter at an urban clinic in Japan: A 4-year observational study

Past clinical data are not currently used to calculate pretest probabilities, as they have not been put into a database in clinical settings. This observational study was designed to determine the initial reasons for utilizing home visits or visits to an outpatient urban clinic in Japan.All family medical clinic outpatients and patients visited by the clinic (total = 11,688) over 1460 days were enrolled.We used a Bayes theorem-based clinical decision support system to analyze codes for initial reason-for-encounter (examination and final diagnosis: pretest probability) and final diagnosis of patients with fever (conditional pretest probability).Total number of reasons-for-encounter: 96,653 (an average of 1.2 reasons per visit). Final diagnosis: 62,273 cases (an average of 0.75 cases per visit). The most common reasons for initial examination were immunizations, physical examinations, and upper respiratory conditions. Regarding the final diagnosis, the combination of physical examinations and acute upper respiratory infections comprised 73.4% of cases. In cases where fever developed, the bulk of the final diagnoses were infectious diseases such as influenza, strep throat, and gastroenteritis of presumed infectious origin. For the elderly, fever often occurred with other health issues such as pneumonia, dementia, constipation, and sleep disturbances, though the cause of the fever remained undetermined in 40% of the cases.The pretest probability changed significantly based on the reason or the combination of reasons for which patients requested a medical examination. Using accumulated data from past diagnoses to modify subsequent subjective diagnoses, individual diagnoses can be improved.

Biologics or tofacitinib for people with rheumatoid arthritis naive to methotrexate: a systematic review and network meta-analysis

BACKGROUND

Biologic disease-modifying anti-rheumatic drugs (biologics) are highly effective in treating rheumatoid arthritis (RA), however there are few head-to-head biologic comparison studies. We performed a systematic review, a standard meta-analysis and a network meta-analysis (NMA) to update the 2009 Cochrane Overview. This review is focused on the adults with RA who are naive to methotrexate (MTX) that is, receiving their first disease-modifying agent.

OBJECTIVES

To compare the benefits and harms of biologics (abatacept, adalimumab, anakinra, certolizumab pegol, etanercept, golimumab, infliximab, rituximab, tocilizumab) and small molecule tofacitinib versus comparator (methotrexate (MTX)/other DMARDs) in people with RA who are naive to methotrexate.

METHODS

In June 2015 we searched for randomized controlled trials (RCTs) in CENTRAL, MEDLINE and Embase; and trials registers. We used standard Cochrane methods. We calculated odds ratios (OR) and mean differences (MD) along with 95% confidence intervals (CI) for traditional meta-analyses and 95% credible intervals (CrI) using a Bayesian mixed treatment comparisons approach for network meta-analysis (NMA). We converted OR to risk ratios (RR) for ease of interpretation. We also present results in absolute measures as risk difference (RD) and number needed to treat for an additional beneficial or harmful outcome (NNTB/H).

MAIN RESULTS

Nineteen RCTs with 6485 participants met inclusion criteria (including five studies from the original 2009 review), and data were available for four TNF biologics (adalimumab (six studies; 1851 participants), etanercept (three studies; 678 participants), golimumab (one study; 637 participants) and infliximab (seven studies; 1363 participants)) and two non-TNF biologics (abatacept (one study; 509 participants) and rituximab (one study; 748 participants)).Less than 50% of the studies were judged to be at low risk of bias for allocation sequence generation, allocation concealment and blinding, 21% were at low risk for selective reporting, 53% had low risk of bias for attrition and 89% had low risk of bias for major baseline imbalance. Three trials used biologic monotherapy, that is, without MTX. There were no trials with placebo-only comparators and no trials of tofacitinib. Trial duration ranged from 6 to 24 months. Half of the trials contained participants with early RA (less than two years' duration) and the other half included participants with established RA (2 to 10 years). Biologic + MTX versus active comparator (MTX (17 trials (6344 participants)/MTX + methylprednisolone 2 trials (141 participants))In traditional meta-analyses, there was moderate-quality evidence downgraded for inconsistency that biologics with MTX were associated with statistically significant and clinically meaningful benefit versus comparator as demonstrated by ACR50 (American College of Rheumatology scale) and RA remission rates. For ACR50, biologics with MTX showed a risk ratio (RR) of 1.40 (95% CI 1.30 to 1.49), absolute difference of 16% (95% CI 13% to 20%) and NNTB = 7 (95% CI 6 to 8). For RA remission rates, biologics with MTX showed a RR of 1.62 (95% CI 1.33 to 1.98), absolute difference of 15% (95% CI 11% to 19%) and NNTB = 5 (95% CI 6 to 7). Biologics with MTX were also associated with a statistically significant, but not clinically meaningful, benefit in physical function (moderate-quality evidence downgraded for inconsistency), with an improvement of HAQ scores of -0.10 (95% CI -0.16 to -0.04 on a 0 to 3 scale), absolute difference -3.3% (95% CI -5.3% to -1.3%) and NNTB = 4 (95% CI 2 to 15).We did not observe evidence of differences between biologics with MTX compared to MTX for radiographic progression (low-quality evidence, downgraded for imprecision and inconsistency) or serious adverse events (moderate-quality evidence, downgraded for imprecision). Based on low-quality evidence, results were inconclusive for withdrawals due to adverse events (RR of 1.32, but 95% confidence interval included possibility of important harm, 0.89 to 1.97). Results for cancer were also inconclusive (Peto OR 0.71, 95% CI 0.38 to 1.33) and downgraded to low-quality evidence for serious imprecision. Biologic without MTX versus active comparator (MTX 3 trials (866 participants)There was no evidence of statistically significant or clinically important differences for ACR50, HAQ, remission, (moderate-quality evidence for these benefits, downgraded for imprecision), withdrawals due to adverse events,and serious adverse events (low-quality evidence for these harms, downgraded for serious imprecision). All studies were for TNF biologic monotherapy and none for non-TNF biologic monotherapy. Radiographic progression was not measured.

AUTHORS' CONCLUSIONS

In MTX-naive RA participants, there was moderate-quality evidence that, compared with MTX alone, biologics with MTX was associated with absolute and relative clinically meaningful benefits in three of the efficacy outcomes (ACR50, HAQ scores, and RA remission rates). A benefit regarding less radiographic progression with biologics with MTX was not evident (low-quality evidence). We found moderate- to low-quality evidence that biologic therapy with MTX was not associated with any higher risk of serious adverse events compared with MTX, but results were inconclusive for withdrawals due to adverse events and cancer to 24 months.TNF biologic monotherapy did not differ statistically significantly or clinically meaningfully from MTX for any of the outcomes (moderate-quality evidence), and no data were available for non-TNF biologic monotherapy.We conclude that biologic with MTX use in MTX-naive populations is beneficial and that there is little/inconclusive evidence of harms. More data are needed for tofacitinib, radiographic progression and harms in this patient population to fully assess comparative efficacy and safety.

On Predicting lung cancer subtypes using 'omic' data from tumor and tumor-adjacent histologically-normal tissue

BACKGROUND

Adenocarcinoma (ADC) and squamous cell carcinoma (SCC) are the most prevalent histological types among lung cancers. Distinguishing between these subtypes is critically important because they have different implications for prognosis and treatment. Normally, histopathological analyses are used to distinguish between the two, where the tissue samples are collected based on small endoscopic samples or needle aspirations. However, the lack of cell architecture in these small tissue samples hampers the process of distinguishing between the two subtypes. Molecular profiling can also be used to discriminate between the two lung cancer subtypes, on condition that the biopsy is composed of at least 50 % of tumor cells. However, for some cases, the tissue composition of a biopsy might be a mix of tumor and tumor-adjacent histologically normal tissue (TAHN). When this happens, a new biopsy is required, with associated cost, risks and discomfort to the patient. To avoid this problem, we hypothesize that a computational method can distinguish between lung cancer subtypes given tumor and TAHN tissue.

METHODS

Using publicly available datasets for gene expression and DNA methylation, we applied four classification tasks, depending on the possible combinations of tumor and TAHN tissue. First, we used a feature selector (ReliefF/Limma) to select relevant variables, which were then used to build a simple naïve Bayes classification model. Then, we evaluated the classification performance of our models by measuring the area under the receiver operating characteristic curve (AUC). Finally, we analyzed the relevance of the selected genes using hierarchical clustering and IPA® software for gene functional analysis.

RESULTS

All Bayesian models achieved high classification performance (AUC > 0.94), which were confirmed by hierarchical cluster analysis. From the genes selected, 25 (93 %) were found to be related to cancer (19 were associated with ADC or SCC), confirming the biological relevance of our method.

CONCLUSIONS

The results from this study confirm that computational methods using tumor and TAHN tissue can serve as a prognostic tool for lung cancer subtype classification. Our study complements results from other studies where TAHN tissue has been used as prognostic tool for prostate cancer. The clinical implications of this finding could greatly benefit lung cancer patients.

Attention to sound improves auditory reliability in audio-tactile spatial optimal integration

The role of attention on multisensory processing is still poorly understood. In particular, it is unclear whether directing attention toward a sensory cue dynamically reweights cue reliability during integration of multiple sensory signals. In this study, we investigated the impact of attention in combining audio-tactile signals in an optimal fashion. We used the Maximum Likelihood Estimation (MLE) model to predict audio-tactile spatial localization on the body surface. We developed a new audio-tactile device composed by several small units, each one consisting of a speaker and a tactile vibrator independently controllable by external software. We tested participants in an attentional and a non-attentional condition. In the attentional experiment, participants performed a dual task paradigm: they were required to evaluate the duration of a sound while performing an audio-tactile spatial task. Three unisensory or multisensory stimuli, conflictual or not conflictual sounds and vibrations arranged along the horizontal axis, were presented sequentially. In the primary task participants had to evaluate in a space bisection task the position of the second stimulus (the probe) with respect to the others (the standards). In the secondary task they had to report occasionally changes in duration of the second auditory stimulus. In the non-attentional task participants had only to perform the primary task (space bisection). Our results showed an enhanced auditory precision (and auditory weights) in the auditory attentional condition with respect to the control non-attentional condition. The results of this study support the idea that modality-specific attention modulates multisensory integration.

Joint spatial modeling to identify shared patterns among chronic related potentially preventable hospitalizations

BACKGROUND

Rates of Potentially Preventable Hospitalizations (PPH) are used to evaluate access of territorially delimited populations to high quality ambulatory care. A common geographic pattern of several PPH would reflect the performance of healthcare providers. This study is aimed at modeling jointly the geographical variation in six chronic PPH conditions in one Spanish Autonomous Community for describing common and discrepant patterns, and to assess the relative weight of the common pattern on each condition.

METHODS

Data on the 39,970 PPH hospital admissions for diabetes short term complications, chronic obstructive pulmonary disease (COPD), congestive heart failure, dehydration, angina admission and adult asthma, between 2007 and 2009 were extracted from the Hospital Discharge Administrative Databases and assigned to one of the 240 Basic Health Zones. Rates and Standardized Hospitalization Ratios per geographic unit were estimated. The spatial analysis was carried out jointly for PPH conditions using Shared Component Models (SCM).

RESULTS

The component shared by the six PPH conditions explained about the 36% of the variability of each PPH condition, ranging from the 25.9 for dehydration to 58.7 for COPD. The geographical pattern found in the latent common component identifies territorial clusters with particularly high risk. The specific risk pattern that each isolated PPH does not share with the common pattern for all six conditions show many non-significant areas for most PPH, but with some exceptions.

CONCLUSIONS

The geographical distribution of the risk of the PPH conditions is captured in a 36% by a unique latent pattern. The SCM modeling may be useful to evaluate healthcare system performance.

Making predictions in a changing world-inference, uncertainty, and learning

To function effectively, brains need to make predictions about their environment based on past experience, i.e., they need to learn about their environment. The algorithms by which learning occurs are of interest to neuroscientists, both in their own right (because they exist in the brain) and as a tool to model participants' incomplete knowledge of task parameters and hence, to better understand their behavior. This review focusses on a particular challenge for learning algorithms-how to match the rate at which they learn to the rate of change in the environment, so that they use as much observed data as possible whilst disregarding irrelevant, old observations. To do this algorithms must evaluate whether the environment is changing. We discuss the concepts of likelihood, priors and transition functions, and how these relate to change detection. We review expected and estimation uncertainty, and how these relate to change detection and learning rate. Finally, we consider the neural correlates of uncertainty and learning. We argue that the neural correlates of uncertainty bear a resemblance to neural systems that are active when agents actively explore their environments, suggesting that the mechanisms by which the rate of learning is set may be subject to top down control (in circumstances when agents actively seek new information) as well as bottom up control (by observations that imply change in the environment).

Spatial proximity and the risk of psychopathology after a terrorist attack

Previous studies concerned with the relation of proximity to the September 11, 2001 terrorist attacks and subsequent psychopathology have produced conflicting results. The goals of this analysis are to assess the appropriateness of using Bayesian hierarchical spatial techniques to answer the question of the role of proximity to a mass trauma as a risk factor for psychopathology. Using a set of individual-level Medicaid data for New York State, and controlling for age, gender, median household income and employment-related exposures, we applied Bayesian hierarchical modeling methods for spatially aggregated data. We found that distance from the World Trade Center site in the post-attack time period was associated with increased risk of anxiety-related diagnoses. In the months following the attack, each 2-mile increment in distance closer to the World Trade Center site was associated with a 7% increase in anxiety-related diagnoses in the population. No similar association was found during a similar time period in the year prior to the attack. We conclude that spatial variables help more fully describe post-terrorism psychiatric risk and may help explain discrepancies in the existing literature about these attacks. These methods hold promise for the characterization of disease risk where spatial patterning of ecologic-level exposures and outcomes merits consideration.

Underlying genetic models of inheritance in established type 2 diabetes associations

For most associations of common single nucleotide polymorphisms (SNPs) with common diseases, the genetic model of inheritance is unknown. The authors extended and applied a Bayesian meta-analysis approach to data from 19 studies on 17 replicated associations with type 2 diabetes. For 13 SNPs, the data fitted very well to an additive model of inheritance for the diabetes risk allele; for 4 SNPs, the data were consistent with either an additive model or a dominant model; and for 2 SNPs, the data were consistent with an additive or recessive model. Results were robust to the use of different priors and after exclusion of data for which index SNPs had been examined indirectly through proxy markers. The Bayesian meta-analysis model yielded point estimates for the genetic effects that were very similar to those previously reported based on fixed- or random-effects models, but uncertainty about several of the effects was substantially larger. The authors also examined the extent of between-study heterogeneity in the genetic model and found generally small between-study deviation values for the genetic model parameter. Heterosis could not be excluded for 4 SNPs. Information on the genetic model of robustly replicated association signals derived from genome-wide association studies may be useful for predictive modeling and for designing biologic and functional experiments.

A fuzzy locally adaptive Bayesian segmentation approach for volume determination in PET

Accurate volume estimation in positron emission tomography (PET) is crucial for different oncology applications. The objective of our study was to develop a new fuzzy locally adaptive Bayesian (FLAB) segmentation for automatic lesion volume delineation. FLAB was compared with a threshold approach as well as the previously proposed fuzzy hidden Markov chains (FHMC) and the fuzzy C-Means (FCM) algorithms. The performance of the algorithms was assessed on acquired datasets of the IEC phantom, covering a range of spherical lesion sizes (10-37 mm), contrast ratios (4:1 and 8:1), noise levels (1, 2, and 5 min acquisitions), and voxel sizes (8 and 64 mm(3)). In addition, the performance of the FLAB model was assessed on realistic nonuniform and nonspherical volumes simulated from patient lesions. Results show that FLAB performs better than the other methodologies, particularly for smaller objects. The volume error was 5%-15% for the different sphere sizes (down to 13 mm), contrast and image qualities considered, with a high reproducibility (variation < 4%). By comparison, the thresholding results were greatly dependent on image contrast and noise, whereas FCM results were less dependent on noise but consistently failed to segment lesions < 2 cm. In addition, FLAB performed consistently better for lesions < 2 cm in comparison to the FHMC algorithm. Finally the FLAB model provided errors less than 10% for nonspherical lesions with inhomogeneous activity distributions. Future developments will concentrate on an extension of FLAB in order to allow the segmentation of separate activity distribution regions within the same functional volume as well as a robustness study with respect to different scanners and reconstruction algorithms.