Israel Medical Association Journal

Background: Potentially preventable readmissions of surgical oncology patients offer opportunities to improve quality of care. Identifying and subsequently addressing remediable causes of readmissions may improve patient-centered care.

Objectives: To identify factors associated with potentially preventable readmissions after index cancer operation.

Methods: The New York State hospital discharge database was used to identify patients undergoing common cancer operations via principal diagnosis and procedure codes between the years 2010 and 2014. The 30-day readmissions were identified and risk factors for potentially preventable readmissions were analyzed using competing risk analysis.

Results: A total of 53,740 cancer surgeries performed for the following tumor types were analyzed: colorectal (CRC) (42%), kidney (22%), liver (2%), lung (25%), ovary (4%), pancreas (4%), and uterine (1%). The 30-day readmission rate was 11.97%, 47% of which were identified as potentially preventable. The most common cause of potentially preventable readmissions was sepsis (48%). Pancreatic cancer had the highest overall readmission rate (22%) and CRC had the highest percentage of potentially preventable readmissions (51%, hazard ratio [HR] 1.42, 95% confidence interval [95%CI] 1.28–1.61). Risk factors associated with preventable readmissions included discharge disposition to a skilled nursing facility (HR 2.22, 95%CI 1.99–2.48) and the need for home healthcare (HR 1.61, 95%CI 1.48–1.75).

Conclusions: Almost half of the 30-day readmissions were potentially preventable and attributed to high rates of sepsis, surgical site infections, dehydration, and electrolyte disorders. These results can be further validated for identifying broad targets for improvement

Background: Patients with severe chronic obstructive pulmonary disease (COPD) experience frequent exacerbations and need to be hospitalized, resulting in an economic and social burden. Although data exist regarding reasons of frequent hospitalizations, there is no data available about the impact on the length of stay (LOS).

Objectives: To characterize the causes of prolonged hospitalizations in COPD patients.

Methods: A retrospective study was conducted of patients who were diagnosed and treated in the pulmonary department for severe COPD exacerbations. All patient demographic data and medical history were collected. Data regarding the disease severity were also collected (including Global Initiative for Obstructive Lung Disease [GOLD] criteria, pulmonologist follow-up, prior hospitalizations, and LOS).

Results: The study comprised 200 patients, average age 69.5 ± 10.8 years, 61% males. Of these patients, 89 (45%) were hospitalized for up to 4 days, 111 (55%) for 5 days or more, and 34 (17%) for more than 7 days. Single patients had longer LOS compared with married patients (48% vs. 34%, P = 0.044). Multivariate analysis showed that the number of prior hospital admissions in the last year was a predictor of LOS (P = 0.038, odds ratio [OR] = 0.807, 95% confidence interval [95%CI] = 0.659–0.988), as well as the use of non-invasive respiratory support by bilevel positive airway pressure (BiPAP) during the hospitalization (P = 0.024, OR = 4.662, 95%CI = 1.229–17.681).

Conclusions: Fewer previous hospitalizations due to COPD exacerbations and the need for non-invasive respiratory support by BiPAP were found as predictors of longer LOS.

Background: Accurate pulse oximetry reading at hospital admission is of utmost importance, mainly for patients presenting with hypoxemia. Nevertheless, there is no accepted or evidence-based protocol for such structured measuring.

Objectives: To devise and assess a structured protocol intended to increase the accuracy of pulse oximetry measurement at hospital admission.

Methods: The authors performed a prospective comparison of protocol-based pulse-oximetry measurement with non-protocol based readings in consecutive patients at hospital admission. They also calculated the relative percentage of improvement for each patient (before and after protocol implementation) as a fraction of the change in peripheral capillary oxygen saturation (SpO₂) from 100%.

Results: A total of 460 patients were recruited during a 6 month period. Implementation of a structured measurement protocol significantly changed saturation values. The SpO₂ values of 24.7% of all study participants increased after protocol implementation (ranging from 1% to 21% increase in SpO₂ values). Among hypoxemic patients (initial SpO₂ < 90%), protocol implementation had a greater impact on final SpO₂ measurements, increasing their median SpO₂ readings by 4% (3–8% interquartile range; P < 0.05). Among this study population, 50% of the cohort improved by 17% of their overall potential and 25% improved by 50% of their overall improvement potential. As for patients presenting with hypoxemia, the median improvement was 31% of their overall SpO₂ potential.

Conclusions: Structured, protocol based pulse-oximetry may improve measurement accuracy and reliability. The authors suggest that implementation of such protocols may improve the management of hypoxemic patients.

Background: Exposure to air pollution in the form of particulate matter smaller than 10 µm in diameter (PM10) has been associated with increased morbidity and mortality. However, since air pollution is correlated with confounding factors that might otherwise affect health, identifying the causal link has proven challenging. 

Objectives: To identify the effect of PM10 on hospital admissions due to respiratory illnesses. 

Methods: We used the Instrumental Variable (IV) methodology to control for confounding factors affecting hospital admissions. Exploiting the timing of sandstorms as an instrumental variable allows for a better estimate of the relationship between PM10 and hospital admissions. Data on PM concentrations and hospital admissions rates were compiled for Israel’s two largest cities, Jerusalem and Tel Aviv, for 2007–2009. We compared our IV estimates to those derived from a Poisson regression, which is commonly used in the existing literature. 

Results: Sandstorms led to an increase of 307 µg/m3 of PM10 concentrations. A 10 µg/m3 increase in PM10 is associated with a 0.8% increase in hospital admissions due to respiratory conditions, using Poisson regression. The same finding was noted using the IV methodology. 

Conclusions: The association between PM10 and hospital admission reflects a primarily causal relationship. Instrumental variable methodology could be applied to analysis of the effect of air pollution on hospital admissions. 

 

Background: In many hospitals a chest X-ray is performed routinely at each patient’s admission. There are scant data regarding its usefulness in contemporary patient populations, which are characterized by patients’ increasing age, severity of illness, and different comorbidities.

Methods: We studied consecutive patients admitted during a 2 month period to a single department of medicine, where hospital policy mandates performing a CXR[1] on admission or soon after. Two senior clinicians who were not involved in the care of these patients assessed the discharge summaries for a clinical indication to perform CXR on admission, as well as its contribution to patient management (major positive, major negative, minor positive, or no contribution). Logistic regression analysis was performed with the SPSS 12 software program.

Results: The study population comprised 675 patients whose mean age was 64.5 ± 17.2 years. Their presenting complaints included chest pain (18%), dyspnea (12%), weakness (10.5%), fever (9%), abdominal pain (8%) and neurologic complaints (7.5%). Physical examination of the chest was normal in 585 (87%) of the cases and abnormal in 87 (13%). Examination of the heart was normal in 518 (77%) and abnormal in 129 (19%). In 19.6% (130 cases) CXR was not performed. Of the 545 CXRs done, 260 (48%) were normal. In only 128 (23.5%) did the admission CXR make a major positive contribution to diagnosis or treatment. In 61 (11.2%) it provided a minor positive contribution and in 153 (28.1%) a major negative contribution. In 184 patients (33.8%) the CXR did not affect either diagnosis or management. It made a major positive contribution to management in patients for whom there was an indication for performing the X-ray (odds ratio 10.3, P < 0.0005) and in those with a relevant finding on physical examination (OR[2] 1.63, P = 0.110). For a major negative contribution of the CXR to management (i.e., ruling out clinically important possibilities), the clinical indication was also very important (OR 72.9, P < 0.005). When patients with either a clinical indication for performing a CXR or an abnormal chest examination were excluded, 329 patients remained (60% of the 545 who had a CXR) in only 12 of them (3.6%) did the routine admission CXR contribute to patient management.

Conclusions: A routine admission CXR has a significant impact on patient management only in those patients in whom there are relevant findings on physical examination or a clear clinical indication for performing the test. There is no need to routinely order CXR on admission to hospital.

[1] CXR = chest X-ray

[2] OR = odds ratio

Background: In the last decade there has been an increase in asthma morbidity. Hospital admission rates for childhood asthma are influenced by the prevalence of asthma and the quality of asthma care.

Objective: To assess trends in hospital admission and readmission rates for childhood asthma in the Jezreel Valley in Israel in the last decade, and to evaluate the possible effect of changes in asthma treatment upon hospitalization for acute asthma during this period.

Methods: All records from pediatric patients from the central hospital in the Jezreel Valley in northeastern Israel over a 10 year period from 1990 through 1999 who were diagnosed as having asthma were thoroughly reviewed and analyzed for admissions, re-admissions, and treatment before and during admissions

Results: There were 1584 admissions, 1208 were first-time admissions and 374 were re-admissions. The number of first-time admissions increased significantly over time (P < 0.0001), with a significant decrease of re-admissions (P < 0.005); this finding was more significant in children under the age of 8 years (P < 0.005). The length of hospital stay decreased significantly from 3.3 days to 2.7 days (P < 0.002). Significant changes in the use of medications included an increase in inhalant glucocorticoids and a decrease in the use of sodium cromoglycate and theophylline. Controller medication use was concomitant with a significant decrease in the re-admission rates.

Conclusions: The increase in the admission rate and the decrease in the rate of re-admissions and the length of hospital stay probably reflect the increase in the prevalence of asthma and changes in its treatment, respectively. It is essential that asthma be recognized as a significant cause of morbidity and that controller medications be administered to decrease the asthma's severity, morbidity, and resultant hospital admissions.
 

Background: The increasing utilization of general internal medicine hospital wards in Israel during the last decade is a source of concern for health policy makers.

Objectives: To report on the distribution of selected main and secondary diagnoses among GIM inpatients, and to estimate the proportion of disorders for which appropriate care in the community will reduce the need for hospital admissions and re-admissions.

Methods: Data from the Health Information and Computer Services of the Israel Ministry of Health (national hospitalization database) for a one year period were analyzed by distribution of diagnostic entities (ICD-9-CM) in GIM and in medical subspecialty wards.

Results: Of the 313,824 discharges from hospital divisions of medicine in 1995, 256,956 (81.9%) were from GIM and 56,868 (18.1%) from specialty wards. Main and secondary discharge diagnoses were available for 188,807 GIM and 35,992 specialty patients. Of all main diagnoses in GIM wards, 27% were coded as "general or systemic symptoms and signs" or as "abnormal laboratory or ill defined manifestations" (ICD-9-CM codes 780-799, 276,277), and heart diseases comprised another 27%. The remaining main diagnoses covered almost all medical conditions. The combined proportion of "ambulatory care sensitive hospital admissions" (bronchial asthma, hypertension, congestive heart failure, chronic obstructive pulmonary disease, diabetes) constituted 12% of all main diagnoses in GIM, and respiratory symptoms or signs comprised another 11%. A by-product of this analysis was an insight into the experience of undergraduate medical students in GIM.

Conclusions: Assuming that 12-75% of admissions for "ambulatory care sensitive disorders" are preventable, an improved review before hospital discharge and a closer outpatient follow-up may reduce the load on GIM wards by 1-17%. This wide range justifies controlled trials to determine the effect of improved community care on hospital utilization. GIM wards offer valuable learning opportunities, but they cannot be a substitute for primary care clinics. The unexplained high proportion of GIM inpatients who were discharged with an unspecified main diagnosis could be detrimental for the accuracy of hospitalization statistics, and justifies investigation by chart audits into physicians' habits of documentation.

__________________________________

GIM= general internal medicine