Frequency and temporal profile of poststroke hyperglycemia using continuous glucose monitoring.

OBJECTIVE: Poststroke hyperglycemia (PSH) is common and has adverse effects on outcome. In this observational study, we aimed to describe the frequency and temporal profile of PSH using a continuous glucose monitoring system (CGMS) in patients with and without diabetes. RESEARCH DESIGN AND METHODS: Fifty-nine patients with acute hemispheric ischemic stroke were prospectively studied with the CGMS, regardless of medication, admission plasma glucose value, and diabetes status. The CGMS records interstitial glucose every 5 min for 72 h. RESULTS: On admission, 36% of patients had preexisting diabetes. At the earliest analyzed time point of 8 h from stroke onset, 50% of nondiabetic subjects and 100% of diabetic patients were hyperglycemic (> or =7 mmol/l). This early-phase hyperglycemia was followed by a decrease in glucose 14-16 h poststroke when only 11% of nondiabetic and 27% of diabetic patients were hyperglycemic. A late hyperglycemic phase 48-88 h poststroke was observed in 27% of nondiabetic and 78% of diabetic patients. Thirty-four percent of nondiabetic and 86% of diabetic patients were hyperglycemic for at least a quarter of the monitoring period. Multivariate regression analysis demonstrated that diabetes, insular cortical ischemia, and increasing age independently predicted higher glucose values. CONCLUSIONS: Poststroke hyperglycemia is common and prolonged despite treatment based on current guidelines. There are early and late hyperglycemic phases in nondiabetic as well as diabetic patients. Treatment protocols with frequent glucose measurement and intensive glucose-lowering therapy for a minimum of 72 h poststroke need to be evaluated.

Tracheostomy decannulation at the Royal Hospital for Sick Children in Glasgow: Predictors of success and failure.

BACKGROUND: Tracheostomy techniques, indications and care are extensively covered in the literature. However, little is written about the process of removing the tracheostomy tube. At the Royal Hospital for Sick Children in Glasgow we use a stepwise ward-based protocol for safe tracheostomy decannulation. Our aim therefore was to review all the paediatric tracheostomy decannulations that we attempted over the last 3 years to evaluate our protocol, to determine our success rate and to see whether any modifications to the protocol are required. METHOD: We reviewed all patients who had undergone ward decannulation between January 2012 and May 2015. We extracted data from clinical records including patient characteristics, indications for tracheostomy, timing of decannulation and success or failure of the process. RESULTS: The 45 children in the study underwent 57 attempts at decannulation during the study period. 25 were male (56%) and 20 were female (44%), and they were aged between 1 day and 16 years 6 months at the time of the original tracheostomy operation. 33 attempts were successful (58%). 10 children had more than one attempt at decannulation. Children were found to fail at every stage of the protocol, with the commonest point of failure being day 2 when the tracheostomy tube was capped. DISCUSSION: We have demonstrated that our current protocol for ward decannulation is effective and safe, and that all five days of the protocol are required.

The safety of hospital stroke care.

OBJECTIVES: To analyze medical errors and adverse events occurring in stroke patients and to provide insights into system or stroke-specific processes that can be modified to reduce the likelihood of error and patient harm. METHODS: We analyzed spontaneously reported errors and adverse events reported within a voluntary and mandatory event reporting system in stroke patients admitted to a 750-bed academic medical center over a 3.5-year period between July 1, 2001, and December 31, 2004. We determined the frequency of near misses and preventable adverse events by event type (medication, adverse clinical, and falls). We performed a central event analysis to determine the most likely cause of preventable adverse events. RESULTS: Of the 1,440 stroke patients admitted during the study period, 173 patients (12.0%) experienced an adverse event that was reported within an event-reporting system. Of the 176 events in 148 patients reported in the voluntary event reporting system, 72 were falls, 62 were medication events, and 42 were adverse clinical events. Of the 28 events in 25 patients reported in the mandatory event-reporting system, all were adverse clinical events and involved patient harm. Of the total 201 unique events (3 events were reported in both systems), 18 were near misses and 183 were adverse events. Of the 183 adverse events, 86 were preventable, 37 were not preventable, and 60 were indeterminate. Preventable adverse events involved drugs and situations commonly seen in the stroke population and occurred in all aspects of care delivery from thrombolytic management to end-of-life care. Of the 86 preventable adverse events, 37% (32/86) were transcription/documentation errors, 23% (20/86) were failure to perform a clinical task, 10% (9/86) were communication/handoff errors between providers, and 10% (9/86) were failed independent checks/calculations. CONCLUSIONS: Adverse events and errors occur frequently in stroke patients. A disease-specific approach to analyzing spontaneously reported events may help close the feedback loop on patient safety and improve the quality of care.

Using administrative data to improve compliance with mandatory state event reporting.

BACKGROUND: The New York Patient Occurrence and Tracking System (NYPORTS) is a mandatory adverse event reporting system that was redesigned in 1998. Analysis of the first full year of its use showed large regional and hospital variation in reporting frequency not due to hospital or case mix differences. In early 2001, New York State mandated that all hospitals conduct retrospective review for unreported adverse incidents for the previous 2 years. Hospitals could submit previously unreported incidents within a defined window without penalty. The hospital used an ICD-9-CM (International Classification of Diseases, Ninth Revision, Clinical Modification) analysis to screen for missed NYPORTS cases, to assist in focusing review resources. METHODS: NYPORTS categories were matched to corresponding combinations of inpatient ICD-9-CM diagnosis and procedure codes. Other variables considered included discharge disposition, primary or secondary coding position, readmissions, and NYPORTS exclusions. RESULTS: Among more than 60,000 discharges in 2 years, 5,500 records were identified for NYPORTS review based on the ICD-9-CM criteria; 211 cases had already been reported through normal reporting processes. Thirteen of the NYPORTS codes had a 30% or greater match rate to the ICD-9-CM codes, with an average "hit rate" of 56%. Five-hundred sixty reviews identified 187 (33.4%) reportable events for the same code the case was being screened for and 26 additional reportable events for a code other than the screening code. NYPORTS categories for procedure and operative-related occurrences had the highest yields. CONCLUSIONS: This retrospective effort helped identify previously unreported occurrences, increase institutional awareness of New York State's mandatory reporting process, and stimulate the redesign of our concurrent detection process.