Elsevier

Resuscitation

Volume 79, Issue 2, November 2008, Pages 241-248
Resuscitation

Clinical paper
Using administrative data to develop a nomogram for individualising risk of unplanned admission to intensive care

https://doi.org/10.1016/j.resuscitation.2008.06.023Get rights and content

Summary

Aim

Although unplanned admissions to the intensive care unit (ICU) are associated with poorer prognoses, there is no published prognostic tool available for predicting this risk in an individual patient. We developed a nomogram for calculating the individualised absolute risk of unplanned ICU admission during a hospital stay.

Method

Hospital administrative data from a large district hospital of consecutive admissions from 1 January 2000 to 31 December 2006 of aged over 14 years was used. Patient data was extracted from 94,482 hospital admissions consisted of demographic and clinical variables, including diagnostic categories, types of admission and time and day of admission. Multivariate logistic regression coefficients were used to develop a predictive nomogram of individual risk to patients admitted to the study hospital of unplanned ICU admission.

Results

A total of 672 incident unplanned ICU admissions were identified over this period. Independent predictors of unplanned ICU admissions included being male, older age, emergency department (ED) admissions, after-hour admissions, weekend admissions and six principal diagnosis groups: fractured femur, acute pancreatitis, liver disease, chronic airway disease, pneumonia and heart failure. The area under the receiver operating characteristic curve was 0.81.

Conclusion

The use of a nomogram to accurately identify at-risk patients using information that is readily available to clinicians has the potential to be a useful tool in reducing unplanned ICU admissions, which in turn may contribute to the reduction of adverse events of patients in the general wards.

Introduction

Patients who are transferred from the general wards to the intensive care unit (ICU) for critical care management, commonly referred to as unanticipated or unplanned ICU admissions,1, 2, 3 have higher mortality rates and poorer prognoses than those who are admitted directly from the emergency department (ED) or the operating rooms.4, 5, 6 Unplanned ICU admission has been used as a surrogate marker of adverse events,2 and is used by the Australian Council of Healthcare Accreditation as a reportable quality indicator.7

Although a number of observational studies have suggested that the Medical Emergency Team (MET) reduced unplanned ICU admissions,3, 8, 9, 10 this was not borne out in a multi-site cluster-randomised controlled trial.11 Whilst not negating the benefits of the MET, findings from this randomised controlled trial gave impetus to explore other strategies of identifying patients who are at risk of unplanned ICU admissions.

One way of objectively identifying at risk patients is the use of vital signs routinely measured as physiological variables, to derive a ‘Patient-At-Risk’ score, as suggested by Goldhill and McNarry.12 However, the authors cautioned that critical thresholds of these scores were likely to be change with different patient cohorts; in addition this algorithm also lacks the ability to individualise risk, thus limiting its usefulness. Added to this, it is also recognised that monitoring and documentation of changes in vital signs are not always adequate in the general wards.11 Taking these into consideration, we sought an alternate method of developing a predictive model, using readily available administrative data to develop a nomogram, which may assist with evidence-based medical decision-making of identifying patients at risk of unplanned ICU admissions.

Nomograms are statistical models designed to provide prognostic information using a combination of variables for an individualised prediction of outcome.13 Nomograms have been shown to be superior to traditional risk grouping systems in identifying at risk patients, with the use of more complex predictive models that maximise individualised prediction.14 The graphical representation of a nomogram enables the magnitude and the direction of each contributing variable to be easily visualised. The use of nomograms has been attributed a change in treatment decision-making process of prostate cancer, which has resulted in improved health outcomes.15 Nomograms are also widely use in osteoporosis research and management, to improve the decision-making of both physicians and patients, in determining the most effective treatment option by taking into account the patient's particular physiological status or disease state.16

Section snippets

Design

A retrospective study design was used to examine the characteristics of unplanned admissions to an intensive care unit over a 7-year period.

Setting

Data for this study were collected at a 370-bed non-teaching hospital in the South Western region of Sydney, Australia that has approximately 22,000 paediatric and adult patients being admitted to the hospital each year. This region is culturally and linguistically diverse, with more than one-third of residents born overseas in a non-English speaking

Results

Six hundred and seventy-two unplanned ICU admissions were identified over the 7-year study period. During this time, the hospital had 94,482 admissions of 54,458 individuals. The admission characteristics overall and in relation to unplanned ICU status are presented in Table 1. The incidence of unplanned ICU admission was greater in men (P < 0.001), and increased with age (P for trend <0.001). Admissions to hospital via the emergency department accounted for 73% of the unplanned ICU admissions.

Discussion

The identification of patients with clinical markers at risk of adverse events such as unplanned ICU admission during a hospital stay has become an important area of research in patient safety.21, 22 For this reason the ability of clinicians to identify patients at increased risk may be valuable to prevent such events from occurring. This study has been able to identify risk factors related to unplanned admission to ICU, and to develop a nomogram for assigning individual risk to patients.

The

Conclusion

This study used readily available administrative data to develop a nomogram, a graphically display of a prediction model that uses numeric scales based on well-established multivariable logistic regression analysis, as a tool for individual risk assessment of unplanned ICU admission. Although independent validation is required, this nomogram has the potential of being a useful tool in assisting clinicians in management-related decisions, as to which patients may require monitoring and follow-up

Conflict of interest

None.

Acknowledgments

Tracy Heath who coordinated the hospital MET system and Jennevie Tan in Clinical Information Department who assisted in retrieving the hospital administrative database.

References (38)

  • M. Sorbellini et al.

    Prognostic nomogram for renal insufficiency after radical or partial nephrectomy

    J Urol

    (2006)
  • E.M.W. van de Garde et al.

    International classification of diseases codes showed modest sensitivity for detecting community-acquired pneumonia

    J Clin Epidemiol

    (2007)
  • N.E. Sharrock

    Fractured femur in the elderly: intensive perioperative care is warranted

    Br J Anaesth

    (2000)
  • M.J. Dacey et al.

    The effect of a rapid response team on major clinical outcome measures in a community hospital

    Crit Care Med

    (2007)
  • G. Haller et al.

    Validity of unplanned admission to an intensive care unit as a measure of patient safety in surgical patients

    Anesthesiology

    (2005)
  • D.R. Goldhill et al.

    Outcome of intensive care patients in a group of British intensive care units

    Crit Care Med

    (1998)
  • J.J. Escarce et al.

    Admission source to the medical intensive care unit predicts hospital death independent of APACHE II score

    JAMA

    (1990)
  • M. Parkhe et al.

    Outcome of emergency department patients with delayed admission to an intensive care unit

    Emerg Med Australas

    (2002)
  • Australian Council for Healthcare Accreditation. Clinical indicators—a user's manual (version 4); 2005....
  • Cited by (25)

    • Evaluation of a digital system to predict unplanned admissions to the intensive care unit: A mixed-methods approach

      2022, Resuscitation Plus
      Citation Excerpt :

      However, EWSs do not account for additional patient risk factors, limiting their sensitivity and generating false alerts.2 Recent studies have shown how scoring systems that use additional variables from the Electronic Patient Record (EPR) (e.g., laboratory results and comorbidities) outperform EWS systems that rely solely on vital signs.3–10 We developed the Hospital Alerting Via Electronic Noticeboard (HAVEN).11

    • Risk factors for unplanned transfer to the intensive care unit after emergency department admission

      2017, American Journal of Emergency Medicine
      Citation Excerpt :

      Prior studies demonstrate that unplanned ICU admissions are associated with worse outcomes and increased mortality [2,3]. This unplanned ICU admission has been used as a surrogate marker of adverse events and is used as a reportable quality indicator by the Australian Council of Healthcare Accreditation [4]. Appropriate disposition of patients presenting to the emergency department (ED) is very important and is likely to improve patient safety and reduce adverse events.

    • Soft tissue oxygenation and risk of mortality (STORM): An early marker of critical illness?

      2015, Journal of Critical Care
      Citation Excerpt :

      The EMR method, not surprisingly, had more discriminatory power but was complex and used many parameters that would not have been available at bedside assessment by RRT. Tam et al [23] performed similar analyses using extensive hemodynamic and biochemical parameters to predict unplanned ICU admissions and derived a model with very good discrimination as well (AUC, 0.81). In essence, both complicated models (Tam and Escobar) have the same lack of timeliness as other robust predictive scoring systems (ie, Acute Physiology and Chronic Health Evaluation, Sequential Organ Failure Assessment, and Multiple Organ Dysfunction Score).

    • Feasibility of using the predisposition, insult/infection, physiological response, and organ dysfunction concept of sepsis to predict the risk of deterioration and unplanned intensive care unit transfer after emergency department admission

      2014, Journal of the Chinese Medical Association
      Citation Excerpt :

      Patients with an unplanned ICU transfer after ED admission have a higher rate of mortality than those who are admitted directly to the ICU from the ED.1–5 Although admissions from the ED account for the majority of unplanned transfers to the ICU,6 strategies to decrease unplanned ICU transfers after ED admission are lacking. Some researchers have tried to determine the risk of unplanned ICU transfer after ED admission7,8; however, these studies were based primarily on administrative data or focused only on certain age groups.

    • Emergency department crowding is associated with decreased quality of care for children with acute asthma

      2011, Annals of Emergency Medicine
      Citation Excerpt :

      Areas for further study would include the quality effect of the surge levels of crowding experienced during the 2009 H1N1 pandemic flu and a reconsideration of the equity models after policy-level changes in health insurance coverage.63,64

    View all citing articles on Scopus

    A Spanish translated version of the summary of this article appears as Appendix in the online version at doi:10.1016/j.resuscitation.2008.06.023.

    View full text