MMMohammed A. Mohammed

Mohammed A. Mohammed

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Mohammed A. Mohammed has a total of 58 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Walking to improve cardiovascular health: a meta-analysis of randomised control trials

    by Elaine M Murtagh, Linda Nichols, Mohammed A Mohammed, Roger Holder, Alan M Nevill, Marie H Murphy


  • Open Access

    Index Blood Tests and National Early Warning Scores within 24 Hours of Emergency Admission Can Predict the Risk of In-Hospital Mortality: A Model Development and Validation Study

    by Mohammed A. Mohammed, Gavin Rudge, Duncan Watson, Gordon Wood, Gary B. Smith, David R. Prytherch, Alan Girling, Andrew Stevens

    We explored the use of routine blood tests and national early warning scores (NEWS) reported within ±24 hours of admission to predict in-hospital mortality in emergency admissions, using empirical decision Tree models because they are intuitive and may ultimately be used to support clinical decision making. A retrospective analysis of adult emergency admissions to a large acute hospital during April 2009 to March 2010 in the West Midlands, England, with a full set of index blood tests results (albumin, creatinine, haemoglobin, potassium, sodium, urea, white cell count and an index NEWS undertaken within ±24 hours of admission). We developed a Tree model by randomly splitting the admissions into a training (50%) and validation dataset (50%) and assessed its accuracy using the concordance (c-) statistic. Emergency admissions (about 30%) did not have a full set of index blood tests and/or NEWS and so were not included in our analysis. There were 23248 emergency admissions with a full set of blood tests and NEWS with an in-hospital mortality of 5.69%. The Tree model identified age, NEWS, albumin, sodium, white cell count and urea as significant (pc-statistic for the training model was 0.864 (95%CI 0.852 to 0.87) and when applied to the testing data set this was 0.853 (95%CI 0.840 to 0.866). An easy to interpret validated risk adjustment Tree model using blood test and NEWS taken within ±24 hours of admission provides good discrimination and offers a novel approach to risk adjustment which may potentially support clinical decision making. Given the nature of the clinical data, the results are likely to be generalisable but further research is required to investigate this promising approach.


  • The findings of the Mid-Staffordshire Inquiry do not uphold the use of hospital standardized mortality ratios as a screening test for 'bad' hospitals

    by M. A. Mohammed, R. Lilford, G. Rudge, R. Holder, A. Stevens

    The Mid-Staffordshire Public Inquiry has published its findings. The initial investigations were triggered by an elevated hospital standardized mortality ratio (HSMR). This shows that the HSMR is being used as a screening test for substandard care; whereby hospitals that fail the test are scrutinized, whilst those that pass the test are not. But screening tests are often misunderstood and misused and so it is prudent to critically examine the HSMR before casting it in the role of a screening test for 'bad' hospitals. A screening test should be valid, have adequate performance characteristics and a clear post-test action plan. The HSMR is not a valid screening test (because the empirical relationship between clinically avoidable mortality and the HSMR is unknown). The HSMR has a poor performance profile (10 of 11 elevated HSMRs would be false alarms and 10 of 11 poorly performing hospitals would escape attention). Crucially, the aim of a post-test investigation into an elevated HSMR is unclear. The use of the HSMR as a screening test for clinically avoidable mortality and thereby substandard care, although well intentioned, is seriously flawed. The findings of the Mid-Staffordshire Public Inquiry have no bearing on this conclusion because a 'bad' hospital cannot uphold a bad screening test. Nonetheless, HSMRs continue to pose a grave public challenge to hospitals, whilst the unsatisfactory nature of the HSMR remains a largely unacknowledged and unchallenged private affair. This asymmetric relationship is inappropriate, unhelpful, costly and potentially harmful. The use of process measures remains a valid way to measure quality of care.


  • Open Access

    A Simple Insightful Approach to Investigating a Hospital Standardised Mortality Ratio: An Illustrative Case-Study

    by Mohammed A. Mohammed, Andrew J. Stevens

    Despite methodological concerns Hospital Standardised Mortality Ratios (HSMRs) are promoted as measures of performance. Hospitals that experience an increase in their HSMR are presented with a serious challenge but with little guidance on how to investigate this complex phenomenon. We illustrate a simple penetrating approach. Retrospective analysis of routinely collected hospital admissions data comparing observed and expected deaths predicted by the Dr Foster Unit case mix adjustment method over three years (n = 74,860 admissions) in Shropshire and Telford NHS Trust Hospital (SaTH) constituting PRH (Princess Royal Hospital) and RSH (Royal Shrewsbury Hospital); whose HSMR increased from 99 in the year 2008/09 to 118 in the year 2009/10. The step up in HSMR was primarily located in PRH (109 to 130 vs. 105 to 118 RSH). Disentangling the HSMR by plotting run charts of observed and expected deaths showed that observed deaths were stable in RSH and PRH but expected deaths, especially at PRH, had fallen. The fall in expected deaths has two possible explanations–genuinely lower risk admissions or that the case-mix adjustment model is underestimating the risk of admissions perhaps because of inadequate clinical coding. There was no evidence that the case-mix profile of admissions had changed but there was considerable evidence that clinical coding process at PRH was producing a lower depth of coding resulting in lower expected mortality. Knowing whether the change (increase/decrease) in HSMR is driven by the numerator or the denominator is a crucial pivotal first step in understanding a given HSMR and so such information should be an integral part of the HSMR reporting methodology.


  • Statistical process control charts for attribute data involving very large sample sizes: a review of problems and solutions

    by Mohammed A Mohammed, Jagdeep S Panesar, David B Laney, Richard Wilson

    The use of statistical process control (SPC) charts in healthcare is increasing. The primary purpose of SPC is to distinguish between common-cause variation which is attributable to the underlying process, and special-cause variation which is extrinsic to the underlying process. This is important because improvement under common-cause variation requires action on the process, whereas special-cause variation merits an investigation to first find the cause. Nonetheless, when dealing with attribute or count data (eg, number of emergency admissions) involving very large sample sizes, traditional SPC charts often produce tight control limits with most of the data points appearing outside the control limits. This can give a false impression of common and special-cause variation, and potentially misguide the user into taking the wrong actions. Given the growing availability of large datasets from routinely collected databases in healthcare, there is a need to present a review of this problem (which arises because traditional attribute charts only consider within-subgroup variation) and its solutions (which consider within and between-subgroup variation), which involve the use of the well-established measurements chart and the more recently developed attribute charts based on Laney's innovative approach. We close by making some suggestions for practice.


  • Open Access

    Factors which influence the length of an out-of-hours telephone consultation in primary care: a retrospective database study

    by Mohammed A Mohammed, Gill Clements, Elaine Edwards, Helen Lester

    Given the increasing use of telephone consultation it is important to determine the factors which influence the length of a telephone consultation.


  • Comparison of four methods for deriving hospital standardised mortality ratios from a single hierarchical logistic regression model

    by Mohammed A Mohammed, Bradley N Manktelow, Timothy P Hofer

    There is interest in deriving case-mix adjusted standardised mortality ratios so that comparisons between healthcare providers, such as hospitals, can be undertaken in the controversial belief that variability in standardised mortality ratios reflects quality of care. Typically standardised mortality ratios are derived using a fixed effects logistic regression model, without a hospital term in the model. This fails to account for the hierarchical structure of the data – patients nested within hospitals – and so a hierarchical logistic regression model is more appropriate. However, four methods have been advocated for deriving standardised mortality ratios from a hierarchical logistic regression model, but their agreement is not known and neither do we know which is to be preferred. We found significant differences between the four types of standardised mortality ratios because they reflect a range of underlying conceptual issues. The most subtle issue is the distinction between asking how an average patient fares in different hospitals versus how patients at a given hospital fare at an average hospital. Since the answers to these questions are not the same and since the choice between these two approaches is not obvious, the extent to which profiling hospitals on mortality can be undertaken safely and reliably, without resolving these methodological issues, remains questionable.


  • Why traditional statistical process control charts for attribute data should be viewed alongside an xmr-chart

    by Mohammed A Mohammed, Peter Worthington

    The use of statistical process control (SPC) charts in healthcare is increasing. The general advice when plotting SPC charts is to begin by selecting the right chart. This advice, in the case of attribute data, may be limiting our insights into the underlying process and consequently be potentially misleading. Given the general lack of awareness that additional insights may be obtained by using more than one SPC chart, there is a need to review this issue and make some recommendations. Under purely common cause variation the control limits on the xmr-chart and traditional attribute charts (eg, p-chart, c-chart, u-chart) will be in close agreement, indicating that the observed variation (xmr-chart) is consistent with the underlying Binomial model (p-chart) or Poisson model (c-chart, u-chart). However, when there is a material difference between the limits from the xmr-chart and the attribute chart then this also constitutes a signal of an underlying systematic special cause of variation. We use one simulation and two case studies to demonstrate these ideas and show the utility of plotting the SPC chart for attribute data alongside an xmr-chart. We conclude that the combined use of attribute charts and xmr-charts, which requires little additional effort, is a useful strategy because it is less likely to mislead us and more likely to give us the insight to do the right thing.


  • Open Access

    Which Is More Useful in Predicting Hospital Mortality -Dichotomised Blood Test Results or Actual Test Values? A Retrospective Study in Two Hospitals

    by Mohammed A. Mohammed, Gavin Rudge, Gordon Wood, Gary Smith, Vishal Nangalia, David Prytherch, Roger Holder, Jim Briggs

    Routine blood tests are an integral part of clinical medicine and in interpreting blood test results clinicians have two broad options. (1) Dichotomise the blood tests into normal/abnormal or (2) use the actual values and overlook the reference values. We refer to these as the “binary” and the “non-binary” strategy respectively. We investigate which strategy is better at predicting the risk of death in hospital based on seven routinely undertaken blood tests (albumin, creatinine, haemoglobin, potassium, sodium, urea, and white blood cell count) using tree models to implement the two strategies. A retrospective database study of emergency admissions to an acute hospital during April 2009 to March 2010, involving 10,050 emergency admissions with routine blood tests undertaken within 24 hours of admission. We compared the area under the Receiver Operating Characteristics (ROC) curve for predicting in-hospital mortality using the binary and non-binary strategy. The mortality rate was 6.98% (701/10050). The mean predicted risk of death in those who died was significantly (p-value <0.0001) lower using the binary strategy (risk = 0.181 95%CI: 0.193 to 0.210) versus the non-binary strategy (risk = 0.222 95%CI: 0.194 to 0.251), representing a risk difference of 28.74 deaths in the deceased patients (n = 701). The binary strategy had a significantly (p-value <0.0001) lower area under the ROC curve of 0.832 (95% CI: 0.819 to 0.845) versus the non-binary strategy (0.853 95% CI: 0.840 to 0.867). Similar results were obtained using data from another hospital. Dichotomising routine blood test results is less accurate in predicting in-hospital mortality than using actual test values because it underestimates the risk of death in patients who died. Further research into the use of actual blood test values in clinical decision making is required especially as the infrastructure to implement this potentially promising strategy already exists in most hospitals.


  • Introducing analysis of means to medical statistics

    by Mohammed A Mohammed, Roger Holder

    Statistical hypothesis testing involving the comparison of three or more means and/or proportions is a frequent undertaking in medical statistics. For comparison of means, analysis of variance is a common choice and for comparison of proportions, χ2 tests are common. However, both these approaches have important limitations which include the need for post hoc testing to identify the unusual group(s) without an integral graphical device to present the final results. These limitations are elegantly overcome by the analysis of means, which is widely used in industrial statistics, and illustrated here using means and proportions.