Search

Multi-state models are useful for modelling disease progression where the state space of the process is used to represent the discrete disease status of subjects. Often, the disease process is only observed at clinical visits, and the schedule of these visits can depend on the disease status of patients. In such situations, the frequency and timing of observations may depend on transition times that are themselves unobserved in an interval-censored setting. There is a potential for bias if we model a disease process with informative observation times as a non-informative observation scheme with pre-specified examination times. In this paper, we develop a joint model for the disease and observation processes to ensure valid inference because the follow-up process may itself contain information about the disease process. The transitions for each subject are modelled using a Markov process, where bivariate subject-specific random effects are used to link the disease and observation models. Inference is based on a Bayesian framework, and we apply our joint model to the analysis of a large study examining functional decline trajectories of palliative care patients.

Transaction costs, liquidity effects, capital limits, and regulatory restrictions reduce arbitrage efficiency and weaken the link between futures prices and their underlying stocks. This study examines how the sampling frequency for return calculation affects the intraday correlation and the lead–lag relationship between the Hong Kong Hang Seng Index futures, the underlying cash index, an exchange-traded fund (ETF) that mimics the index, and each component stock. The paper uses firm bid and offer quotes to eliminate the negative bias in measuring the correlation between the futures and the less liquid stocks. The sampling frequency significantly affects the futures-stock dynamic relationship, particularly for small-cap stocks. The overall results indicate that market participants trade ETF and the largest index's constituent stocks to capitalize on index arbitrage opportunities. The lag of the cash index is greatest against futures price movement, a reflection of the adverse impact of stale prices on the informational role of stock indexes. © 2015 Wiley Periodicals, Inc. Jrl Fut Mark 35:939–952, 2015

BACKGROUND Institutions have tried to replace the use of numeric pagers for clinical communication by implementing health information technology (HIT) solutions. However, failing to account for the sociotechnical aspects of HIT or the interplay of technology with existing clinical workflow, culture, and social interactions may create other unintended consequences. OBJECTIVE To evaluate a Web-based messaging system that allows asynchronous communication between health providers and identify the unintended consequences associated with implementing such technology. DESIGN Intervention—a Web-based messaging system at the University Health Network to replace numeric paging practices in May 2010. The system facilitated clinical communication on the medical wards for coordinating patient care. Study design—pre-post mixed methods utilizing both quantitative and qualitative measures. PARTICIPANTS Five residents, 8 nurses, 2 pharmacists, and 2 social workers were interviewed. Pre-post interruption—15 residents from 5 clinical teams in both periods. MEASUREMENTS The study compared the type of messages sent to physicians before and after implementation of the Web-based messaging system; a constant comparative analysis of semistructured interviews was used to generate key themes related to unintended consequences. RESULTS Interruptions increased 233%, from 3 pages received per resident per day pre-implementation to 10 messages received per resident per day post-implementation. Key themes relating to unintended consequences that emerged from the interviews included increase in interruptions, accountability, and tactics to improve personal productivity. CONCLUSIONS Meaningful improvements in clinical communication can occur but require more than just replacing pagers. Introducing HIT without addressing the sociotechnical aspects of HIT that underlie clinical communication can lead to unintended consequences. Journal of Hospital Medicine 2013;8:137–143. © 2013 Society of Hospital Medicine

This study describes the significant correlation between the Braden Scale (BS) and the Palliative Performance Scale (PPS) in patients with advanced illness that has not been previously reported. The analysis was based on a prospective sequential case series of 664 patients suffering from advanced illness who were referred to a regional palliative medicine programme in Toronto, Canada. Baseline BS and PPS scores assessed within 24 hours of referral were considered for analysis. After controlling for age, gender, consult site and diagnosis (cancer versus non cancer), we observed a significant positive correlation between baseline PPS and BS scores (r ¼ 0� 885, P , 0� 001). These findings suggest that for patients with advanced illness where BS is not routinely used, PPS could be considered as a proxy for pressure ulcer risk assessment.