Reducing surgical mortality in Scotland by use of the WHO Surgical Safety Checklist.

BACKGROUND: The WHO Surgical Safety Checklist has been implemented widely since its launch in 2008. It was introduced in Scotland as part of the Scottish Patient Safety Programme (SPSP) between 2008 and 2010, and is now integral to surgical practice. Its influence on outcomes, when analysed at a population level, remains unclear. METHODS: This was a population cohort study. All admissions to any acute hospital in Scotland between 2000 and 2014 were included. Standardized differences were used to estimate the balance of demographics over time, after which interrupted time-series (segmented regression) analyses were performed. Data were obtained from the Information Services Division, Scotland. RESULTS: There were 12 667 926 hospital admissions, of which 6 839 736 had a surgical procedure. Amongst the surgical cohort, the inpatient mortality rate in 2000 was 0·76 (95 per cent c.i. 0·68 to 0·84) per cent, and in 2014 it was 0·46 (0·42 to 0·50) per cent. The checklist was associated with a 36·6 (95 per cent c.i. -55·2 to -17·9) per cent relative reduction in mortality (P < 0·001). Mortality rates before implementation were decreasing by 0·003 (95 per cent c.i. -0·017 to +0·012) per cent per year; annual decreases of 0·069 (-0·092 to -0·046) per cent were seen during, and 0·019 (-0·038 to +0·001) per cent after, implementation. No such improvement trends were seen in the non-surgical cohort over this time frame. CONCLUSION: Since the implementation of the checklist, as part of an overall national safety strategy, there has been a reduction in perioperative mortality.

Minimally invasive surgery and its impact on 30-day postoperative complications, unplanned readmissions and mortality.

BACKGROUND: A critical appraisal of the benefits of minimally invasive surgery (MIS) is needed, but is lacking. This study examined the associations between MIS and 30-day postoperative outcomes including complications graded according to the Clavien-Dindo classification, unplanned readmissions, hospital stay and mortality for five common surgical procedures. METHODS: Patients undergoing appendicectomy, colectomy, inguinal hernia repair, hysterectomy and prostatectomy were identified in the American College of Surgeons National Surgical Quality Improvement Program database. Non-parsimonious propensity score methods were used to construct procedure-specific matched-pair cohorts that reduced baseline differences between patients who underwent MIS and those who did not. Bonferroni correction for multiple comparisons was applied and P < 0·006 was considered significant. RESULTS: Of the 532 287 patients identified, 53·8 per cent underwent MIS. Propensity score matching yielded an overall sample of 327 736 patients (appendicectomy 46 688, colectomy 152 114, inguinal hernia repair 59 066, hysterectomy 59 066, prostatectomy 10 802). Within the procedure-specific matched pairs, MIS was associated with significantly lower odds of Clavien-Dindo grade I-II, III and IV complications (P ≤ 0·004), unplanned readmissions (P < 0·001) and reduced hospital stay (P < 0·001) in four of the five procedures studied, with the exception of inguinal hernia repair. The odds of death were lower in patients undergoing MIS colectomy (P < 0·001), hysterectomy (P = 0·002) and appendicectomy (P = 0·002). CONCLUSION: MIS was associated with significantly fewer 30-day postoperative complications, unplanned readmissions and deaths, as well as shorter hospital stay, in patients undergoing colectomy, prostatectomy, hysterectomy or appendicectomy. No benefits were noted for inguinal hernia repair.

Physician Perceptions and Beliefs about Generating and Providing a Clinical Summary of the Office Visit.

BACKGROUND: A core measure of the meaningful use of EHR incentive program is the generation and provision of the clinical summary of the office visit, or the after visit summary (AVS), to patients. However, little research has been conducted on physician perceptions and beliefs about the AVS. OBJECTIVES: Evaluate physician perceptions and beliefs about the AVS and the effect of the AVS on workload, patient outcomes, and the care the physician delivers. METHODS: A cross-sectional online survey of physicians at two academic medical centers (AMCs) in the northeast who are participating in the meaningful use EHR incentive program. RESULTS: Of the 1 795 physicians at both AMCs participating in the incentive program, 853 completed the survey for a response rate of 47.5%. Eighty percent of the respondents reported that the AVS was easy (very easy or quite easy or somewhat easy) to generate and provide to patients. Nonetheless, more than three-fourths of the respondents reported a negative effect of generating and providing the AVS on workload of office staff (78%) and workload of physicians (76%). Primary care physicians had more positive beliefs about the effect of the AVS on patient outcomes than specialists (p<0.001) and also had more positive beliefs about the effect of the AVS on the care they delivered than specialists (p<0.001). CONCLUSIONS: Achieving the core meaningful use measure of generating and providing the AVS was easy for physicians but it did not necessarily translate into positive beliefs about the effect of the AVS on patient outcomes or the care the physician delivered. Physicians also had negative beliefs about the effect of the AVS on workload. To promote positive beliefs among physicians around the AVS, organizations should obtain physician input into the design and implementation of the AVS and develop strategies to mitigate its negative impacts on workload.

Physician beliefs about the impact of meaningful use of the EHR: a cross-sectional study.

BACKGROUND: As adoption and use of electronic health records (EHRs) grows in the United States, there is a growing need in the field of applied clinical informatics to evaluate physician perceptions and beliefs about the impact of EHRs. The meaningful use of EHR incentive program provides a suitable context to examine physician beliefs about the impact of EHRs. OBJECTIVE: Contribute to the sparse literature on physician beliefs about the impact of EHRs in areas such as quality of care, effectiveness of care, and delivery of care. METHODS: A cross-sectional online survey of physicians at two academic medical centers (AMCs) in the northeast who were preparing to qualify for the meaningful use of EHR incentive program. RESULTS: Of the 1,797 physicians at both AMCs who were preparing to qualify for the incentive program, 967 completed the survey for an overall response rate of 54%. Only 23% and 27% of physicians agreed or strongly agreed that meaningful use of the EHR will help them improve the care they personally deliver and improve quality of care respectively. Physician specialty was significantly associated with beliefs; e.g., 35% of primary care physicians agreed or strongly agreed that meaningful use will improve quality of care compared to 26% of medical specialists and 21% of surgical specialists (p=0.009). Satisfaction with outpatient EHR was also significantly related to all belief items. CONCLUSIONS: Only about a quarter of physicians in our study responded positively that meaningful use of the EHR will improve quality of care and the care they personally provide. These findings are similar to and extend findings from qualitative studies about negative perceptions that physicians hold about the impact of EHRs. Factors outside of the regulatory context, such as physician beliefs, need to be considered in the implementation of the meaningful use of the EHR incentive program.

Mortality and drug exposure in a 5-year cohort of patients with chronic liver disease.

BACKGROUND: Chronic liver diseases are common in the general population. Drug treatment in this group may be challenging, as many drugs are hepatically metabolised and hepatotoxic. OBJECTIVES: We aimed to assess the mortality of patients with chronic liver disease according to specific drug exposures and the three laboratory parameters creatinine, bilirubin and International Normalised Ratio (INR). METHODS: We conducted a multicentre, 5-year retrospective cohort study in two tertiary university referral hospitals and a secondary referral hospital, using a research database to evaluate the crude and adjusted mortality. RESULTS: Of 1159362 individual patients 1.7% (n = 20158) had chronic liver disease and in this group 36.8% had unspecified chronic non-alcoholic liver disease, 30.1% chronic hepatitis C and 11.9% cirrhosis of the liver. 8.4% of patients presented a diagnosis associated with alcohol. The 4-year survival rates were significantly higher in the group with the most normal laboratory values (94.3%) versus 34.5% in the group with elevated parameters (p <0.001). Overall, drug exposure was not associated with higher mortality; in adjusted multivariate analysis the hazard ratio for anti-cancer drugs was 2.69 (95% CI 1.32-5.46). Of individual drugs, mortality hazard ratios for amiodarone, morphine oral, acetazolamide, sirolimus and lamivudine were 2.46 (95% CI 1.68-3.61), 2.26 (95% CI 1.78-2.86), 2.10 (95% CI 1.19-3.70), 1.81 (95% CI 1.02-3.21) and 1.72 (95% CI 1.17-2.53) respectively. CONCLUSIONS: Drug exposure in general was not associated with higher mortality except for a few categories. Mortality in patients with chronic liver disease was high and is associated with simple laboratory values.

Bias in estimating association parameters for longitudinal binary responses with drop-outs.

This paper considers the impact of bias in the estimation of the association parameters for longitudinal binary responses when there are drop-outs. A number of different estimating equation approaches are considered for the case where drop-out cannot be assumed to be a completely random process. In particular, standard generalized estimating equations (GEE), GEE based on conditional residuals, GEE based on multivariate normal estimating equations for the covariance matrix, and second-order estimating equations (GEE2) are examined. These different GEE estimators are compared in terms of finite sample and asymptotic bias under a variety of drop-out processes. Finally, the relationship between bias in the estimation of the association parameters and bias in the estimation of the mean parameters is explored.

Non-invasive assessment of rejection in pediatric transplant patients: serologic and echocardiographic prediction of biopsy-proven myocardial rejection.

BACKGROUND: Cardiac allograft rejection is a multifocal immune process that is currently assessed using biopsy-guided histologic classification systems (International Society for Heart and Lung Transplantation). Cardiac troponin T and I are established serologic markers of global myocyte damage. The use of load-independent measures of contractility have also been shown to accurately assess the presence of ventricular dysfunction. Little is known about their utility in accurately predicting rejection in the pediatric age group. We undertook the present study to compare rejection grade with echocardiographic and serologic estimates of transplant rejection-related myocardial damage. METHODS: We compared histologic rejection grades (0 to 4) with patient characteristics, echocardiographic measurements, catheterization measurements, and biochemical markers for 86 evaluations in 37 transplant recipients at Children's Hospital. RESULTS: In univariate analyses, biopsy scores correlated (p < 0.05) inversely with left ventricular systolic function (shortening fraction) and contractility (stress velocity index, SVI), and directly with mitral E-wave amplitude. In multivariate analyses, lower contractility and higher mitral E-wave amplitude remained significantly (p < or = 0.01) associated with rejection (SVI, p = 0.002, odds ratio = 0.393; E wave, p = 0.0002, odds ratio = 228). Most rejection episodes were associated with elevation of biochemical markers of myocardial injury. Although troponin I was weakly associated with differences between rejection grades (p = 0.034), troponin T, creatine kinase-MB fraction, and C-reactive protein did not differ with biopsy-rejection scores. Serum markers had a poor predictive capacity for biopsy-detected rejection. Troponin T and I did correlate with increased left ventricular wall thickness and mass. CONCLUSION: Progressively depressed left ventricular contractility and diastolic function are found with worsening pediatric heart transplant rejection-biopsy score; however, sensitive and specific serum markers do not correspond to the degree of active myocardial injury. The use of echocardiographic measures of contractility is associated with a specificity of 91.8% but low sensitivity of 66.7%. Overall we found poor concordance between serum markers and grade of rejection. It is unclear whether myocardial injury as assessed by serum markers, echocardiography, or histologic scoring is more important for assessment of acute rejection or long-term outcome, but it does not appear that serum and tissue markers of rejection can be used interchangeably.

GEE with Gaussian estimation of the correlations when data are incomplete.

This paper considers a modification of generalized estimating equations (GEE) for handling missing binary response data. The proposed method uses Gaussian estimation of the correlation parameters, i.e., the estimating function that yields an estimate of the correlation parameters is obtained from the multivariate normal likelihood. The proposed method yields consistent estimates of the regression parameters when data are missing completely at random (MCAR). However, when data are missing at random (MAR), consistency may not hold. In a simulation study with repeated binary outcomes that are missing at random, the magnitude of the potential bias that can arise is examined. The results of the simulation study indicate that, when the working correlation matrix is correctly specified, the bias is almost negligible for the modified GEE. In the simulation study, the proposed modification of GEE is also compared to the standard GEE, multiple imputation, and weighted estimating equations approaches. Finally, the proposed method is illustrated using data from a longitudinal clinical trial comparing two therapeutic treatments, zidovudine (AZT) and didanosine (ddI), in patients with HIV.

Estimation with correlated censored survival data with missing covariates.

Incomplete covariate data are a common occurrence in studies in which the outcome is survival time. Further, studies in the health sciences often give rise to correlated, possibly censored, survival data. With no missing covariate data, if the marginal distributions of the correlated survival times follow a given parametric model, then the estimates using the maximum likelihood estimating equations, naively treating the correlated survival times as independent, give consistent estimates of the relative risk parameters Lipsitz et al. 1994 50, 842-846. Now, suppose that some observations within a cluster have some missing covariates. We show in this paper that if one naively treats observations within a cluster as independent, that one can still use the maximum likelihood estimating equations to obtain consistent estimates of the relative risk parameters. This method requires the estimation of the parameters of the distribution of the covariates. We present results from a clinical trial Lipsitz and Ibrahim (1996b) 2, 5-14 with five covariates, four of which have some missing values. In the trial, the clusters are the hospitals in which the patients were treated.

Modeling kappa for measuring dependent categorical agreement data.

A method for analysing dependent agreement data with categorical responses is proposed. A generalized estimating equation approach is developed with two sets of equations. The first set models the marginal distribution of categorical ratings, and the second set models the pairwise association of ratings with the kappa coefficient (kappa) as a metric. Covariates can be incorporated into both sets of equations. This approach is compared with a latent variable model that assumes an underlying multivariate normal distribution in which the intraclass correlation coefficient is used as a measure of association. Examples are from a cervical ectopy study and the National Heart, Lung, and Blood Institute Veteran Twin Study.

Non-ignorable missing covariates in generalized linear models.

We propose a likelihood method for estimating parameters in generalized linear models with missing covariates and a non-ignorable missing data mechanism. In this paper, we focus on one missing covariate. We use a logistic model for the probability that the covariate is missing, and allow this probability to depend on the incomplete covariate. We allow the covariates, including the incomplete covariate, to be either categorical or continuous. We propose an EM algorithm in this case. For a missing categorical covariate, we derive a closed form expression for the E- and M-steps of the EM algorithm for obtaining the maximum likelihood estimates (MLEs). For a missing continuous covariate, we use a Monte Carlo version of the EM algorithm to obtain the MLEs via the Gibbs sampler. The methodology is illustrated using an example from a breast cancer clinical trial in which time to disease progression is the outcome, and the incomplete covariate is a quality of life physical well-being score taken after the start of therapy. This score may be missing because the patients are sicker, so this covariate could be non-ignorably missing.

Missing data perspectives of the fluvoxamine data set: a review.

Fitting models to incomplete categorical data requires more care than fitting models to the complete data counterparts, not only in the setting of missing data that are non-randomly missing, but even in the familiar missing at random setting. Various aspects of this point of view have been considered in the literature. We review it using data from a multi-centre trial on the relief of psychiatric symptoms. First, it is shown how the usual expected information matrix (referred to as naive information) is biased even under a missing at random mechanism. Second, issues that arise under non-random missingness assumptions are illustrated. It is argued that at least some of these problems can be avoided using contextual information.

GEECAT and GEEGOR: computer programs for the analysis of correlated categorical response data.

GEECAT and GEEGOR are two user-friendly SAS macros for the analysis of clustered, correlated categorical response data. Both programs implement methodology which extend the generalized estimating equation (GEE) approach of Liang and Zeger (Biometrika 73 (1986) 13-22). GEECAT and GEEGOR both use a first set of estimating equations to model the marginal response. With GEECAT, either correlated nominal or ordered categorical response data can be analyzed. The program GEEGOR employs a second set of estimating equations to model the association of ordered categorical responses within a cluster using the global odds ratio as a measure of association. The programs run on both mainframe computers and microcomputers. Examples are provided to illustrate the features of both programs.

Using missing data methods in genetic studies with missing mutation status.

Because of current techniques of determining gene mutation, investigators are now interested in estimating the odds ratio between genetic status (mutation, no mutation) and an outcome variable such as disease cell type (A, B). In this paper we consider the mutation of the RAS genetic family. To determine if the genes have mutated, investigators look at five specific locations on the RAS gene. RAS mutated is a mutation in at least one of the five gene locations and RAS non-mutated is no mutation in any of the five locations. Owing to limited time and financial resources, one cannot obtain a complete genetic evaluation of all five locations on the gene for all patients. We propose the use of maximum likelihood (ML) with a 2(6) multinomial distribution formed by cross-classifying the binary mutation status at five locations by binary disease cell type. This ML method includes all patients regardless of completeness of data, treats the locations not evaluated as missing data, and uses the EM algorithm to estimate the odds ratio between genetic mutation status and the disease type. We compare the ML method to complete case estimates, and a method used by clinical investigators, which excludes patients with data on less than five locations who have no mutations on these sites.

Goodness-of-fit for GEE: an example with mental health service utilization.

Suppose we use generalized estimating equations to estimate a marginal regression model for repeated binary observations. There are no established summary statistics available for assessing the adequacy of the fitted model. In this paper we propose a goodness-of-fit test statistic which has an approximate chi-squared distribution when we have specified the model correctly. The proposed statistic can be viewed as an extension of the Hosmer and Lemeshow goodness-of-fit statistic for ordinary logistic regression to marginal regression models for repeated binary responses. We illustrate the methods using data from a study of mental health service utilization by children. The repeated responses are a set of binary measures of service use. We fit a marginal logistic regression model to the data using generalized estimating equations, and we apply the proposed goodness-of-fit statistic to assess the adequacy of the fitted model.

Selection models and pattern-mixture models for incomplete data with covariates.

Most models for incomplete data are formulated within the selection model framework. This paper studies similarities and differences of modeling incomplete data within both selection and pattern-mixture settings. The focus is on missing at random mechanisms and on categorical data. Point and interval estimation is discussed. A comparison of both approaches is done on side effects in a psychiatric study.

Likelihood methods for incomplete longitudinal binary responses with incomplete categorical covariates.

We consider longitudinal studies in which the outcome observed over time is binary and the covariates of interest are categorical. With no missing responses or covariates, one specifies a multinomial model for the responses given the covariates and uses maximum likelihood to estimate the parameters. Unfortunately, incomplete data in the responses and covariates are a common occurrence in longitudinal studies. Here we assume the missing data are missing at random (Rubin, 1976, Biometrika 63, 581-592). Since all of the missing data (responses and covariates) are categorical, a useful technique for obtaining maximum likelihood parameter estimates is the EM algorithm by the method of weights proposed in Ibrahim (1990, Journal of the American Statistical Association 85, 765-769). In using the EM algorithm with missing responses and covariates, one specifies the joint distribution of the responses and covariates. Here we consider the parameters of the covariate distribution as a nuisance. In data sets where the percentage of missing data is high, the estimates of the nuisance parameters can lead to highly unstable estimates of the parameters of interest. We propose a conditional model for the covariate distribution that has several modeling advantages for the EM algorithm and provides a reduction in the number of nuisance parameters, thus providing more stable estimates in finite samples.

A global goodness-of-fit statistic for Cox regression models.

In this paper, a global goodness-of-fit test statistic for a Cox regression model, which has an approximate chi-squared distribution when the model has been correctly specified, is proposed. Our goodness-of-fit statistic is global and has power to detect if interactions or higher order powers of covariates in the model are needed. The proposed statistic is similar to the Hosmer and Lemeshow (1980, Communications in Statistics A10, 1043-1069) goodness-of-fit statistic for binary data as well as Schoenfeld's (1980, Biometrika 67, 145-153) statistic for the Cox model. The methods are illustrated using data from a Mayo Clinic trial in primary billiary cirrhosis of the liver (Fleming and Harrington, 1991, Counting Processes and Survival Analysis), in which the outcome is the time until liver transplantation or death. The are 17 possible covariates. Two Cox proportional hazards models are fit to the data, and the proposed goodness-of-fit statistic is applied to the fitted models.

Monte Carlo EM for missing covariates in parametric regression models.

We propose a method for estimating parameters for general parametric regression models with an arbitrary number of missing covariates. We allow any pattern of missing data and assume that the missing data mechanism is ignorable throughout. When the missing covariates are categorical, a useful technique for obtaining parameter estimates is the EM algorithm by the method of weights proposed in Ibrahim (1990, Journal of the American Statistical Association 85, 765-769). We extend this method to continuous or mixed categorical and continuous covariates, and for arbitrary parametric regression models, by adapting a Monte Carlo version of the EM algorithm as discussed by Wei and Tanner (1990, Journal of the American Statistical Association 85, 699-704). In addition, we discuss the Gibbs sampler for sampling from the conditional distribution of the missing covariates given the observed data and show that the appropriate complete conditionals are log-concave. The log-concavity property of the conditional distributions will facilitate a straightforward implementation of the Gibbs sampler via the adaptive rejection algorithm of Gilks and Wild (1992, Applied Statistics 41, 337-348). We assume the model for the response given the covariates is an arbitrary parametric regression model, such as a generalized linear model, a parametric survival model, or a nonlinear model. We model the marginal distribution of the covariates as a product of one-dimensional conditional distributions. This allows us a great deal of flexibility in modeling the distribution of the covariates and reduces the number of nuisance parameters that are introduced in the E-step. We present examples involving both simulated and real data.

Using ecological data to estimate a regression model for individual data: the association between arsenic in drinking water and incidence of skin cancer.

In ecologic studies, participants are studied by groups, and the exposure status of each group is usually represented by a single indicator, mostly the mean exposure. In this paper, we propose using multiple variables derived from dummy variables at the individual level to describe the exposure. An analysis of the association between arsenic in drinking water and skin cancer was used as an example. Well water arsenic levels and skin cancer incidence from 1980 to 1987 were assessed for 243 townships in Taiwan. We first analyzed the data using the mean arsenic concentration in each township as the only exposure variable. The second analysis used multiple variables to describe arsenic exposure; each variable denoted the percentage of wells with arsenic levels within a specific range in each township. Although the first approach did not identify associations between arsenic levels and skin cancer, the multiple-variable approach identifies a positive association at the highest arsenic exposure category (>0.64 mg/L) in both men and women. Therefore, using multiple variables to describe an exposure in ecologic studies may facilitate a better description of the exposure status and thereby lead to more accurate risk assessment, especially when the dose-response relationship is not linear.