Skip Ribbon Commands
Skip to main content
SharePoint

PublicComment

 Feedback

If you have any comments or questions on this page, please email us at:
 

METs for adults

Question Type:
Intervention
Full Question:
Among adults who are at risk for cardiac or respiratory arrest in the hospital (P), does EWSS/response teams/MET systems (I), compared with no such responses (C), change survival to hospital discharge, In-hospital incidence of cardiac/respiratory arrest, survival to hospital discharge with good neurological outcome (O)?
Consensus on Science:
For the critical outcome of survival to hospital discharge, we have found low-quality evidence (downgraded for risk of bias and inconsistency) from 2 RCTs(Priestley 2004, 1398; Hillman 2005, 2091) and very-low-quality evidence (downgraded for risk of bias, inconsistency, and indirectness) from 30 non-RCTs.(Bristow 2000, 236; Buist 2002, 387; Bellomo 2003, 283; Subbe 2003, 797; Kenward 2004, 257; Dacey 2007, 2076; Baxter 2008, 223; Chan 2008, 2506; Rothschild 2008, 417; Campello 2009, 3054; Snyder 2009, 834; Vazquez 2009, 449; Konrad 2010, 100; Lighthall 2010, 679; Santamaria 2010, 445; Beitler 2011, R269; Hayani 2011, 1138; Jones 2011, 83; Laurens 2011, 707; Lim 2011, 373; Moon 2011, 150; Patel 2011, 1455; Sarani 2011, 415; Shah 2011, 1361; Howell 2012, 2562; Rothberg 2012, 98; Sabahi 2012, 270; Scherr 2012, 32; Simmes 2012, 20; Al-Qahtani 2013, 506; Chen 2014, 167; Salvatierra 2014, 2001) Of the 2 RCTs, one demonstrated no significant difference between control hospitals (functioned as usual) and intervention hospitals (introduced a MET team) for both unadjusted (P=0.564; Diff, −0.093; 95% CI, −0.423 to 0.237) and adjusted (P=0.752; OR, 1.03; 95% CI, 0.84–1.28) survival.(Hillman 2005, 2091) The other study demonstrated a significant difference between control wards and intervention wards (introduction of a critical care outreach service) with all patients (OR, 0.70; 95% CI, 0.50–0.97), and matched randomized patients (OR, 0.52; 95% CI, 0.32–0.85).(Priestley 2004, 1398) Of the 33 nonrandomized studies reporting mortality, no studies reported statistically significant worse outcomes for the intervention; 15 studies with no adjustment demonstrated no significant improvement(Subbe 2003, 797; Kenward 2004, 257; Baxter 2008, 223; Rothschild 2008, 417; Campello 2009, 3054; Snyder 2009, 834; Vazquez 2009, 449; Hayani 2011, 1138; Jones 2011, 83; Lim 2011, 373; Patel 2011, 1455; Shah 2011, 1361; Rothberg 2012, 98; Scherr 2012, 32; Simmes 2012, 20); 6 studies with no adjustment demonstrated significant improvement(Buist 2002, 387; Bellomo 2003, 283; Laurens 2011, 707; Moon 2011, 150; Sabahi 2012, 270; Al-Qahtani 2013, 506); 1 study with no adjustment reported on rates, which improved with MET, but did not report on significance(Dacey 2007, 2076); 1 study with no adjustment demonstrated significant improvement for medical patients but not surgical patients (combined significance not reported)(Sarani 2011, 415); 4 studies with adjustment demonstrated significant improvement both before and after adjustment(Konrad 2010, 100; Beitler 2011, R269; Chen 2014, 167); 2 studies with adjustment demonstrated no significant improvement both before and after adjustment(Bristow 2000, 236; Chan 2008, 2506); 2 studies with adjustment demonstrated significant improvement before adjustment but not after adjustment(Lighthall 2010, 679; Salvatierra 2014, 2001); 1 study with adjustment demonstrated significant improvement before adjustment but not after adjustment(Todd 1998, 364); 1 study that reported on both unexpected mortality and overall mortality showed significant improvement both before and after adjustment for unexpected mortality but no significant improvement both before and after adjustment for overall mortality(Santamaria 2010, 445); and 1 before-after study that presented “after” data for unexpected mortality in 3 separate time bands demonstrated significant improvement in time band 3 before adjustment and in time bands 2 and 3 after adjustment.(Howell 2012, 2562) The heterogeneous nature of the studies prevents pooling of data; however, there is a suggestion of improved hospital survival in those hospitals that introduce a MET service, and a suggestion of a dose-response effect, with higher-intensity systems (eg, higher MET calling rates, senior medical staff on MET teams) being more effective. For the critical outcome of in-hospital incidence of cardiac/respiratory arrest, we found low-quality evidence (downgraded for risk of bias and indirectness) from 1 RCT(Hillman 2005, 2091) and very-low-quality evidence (downgraded for risk of bias, inconsistency, and indirectness) from 31 further non-RCTs.(Bristow 2000, 236; Buist 2002, 387; Bellomo 2003, 283; Subbe 2003, 797; DeVita 2004, 251; Kenward 2004, 257; Dacey 2007, 2076; Offner 2007, 1223; Baxter 2008, 223; Benson 2008, 743; Rothschild 2008, 417; Campello 2009, 3054; Moldenhauer 2009, 164; Vazquez 2009, 449; Konrad 2010, 100; Lighthall 2010, 679; Santamaria 2010, 445; Beitler 2011, R269; Laurens 2011, 707; Lim 2011, 373; Moon 2011, 150; Sarani 2011, 415; Shah 2011, 1361; Rothberg 2012, 98; Sabahi 2012, 270; Scherr 2012, 32; Simmes 2012, 20; Al-Qahtani 2013, 506; Chen 2014, 167) For the 1 RCT,(Hillman 2005, 2091) no significant difference between control hospitals and intervention hospitals, both unadjusted (P=0.306; Diff, −0.208; 95% CI, −0.620 to 0.204) and adjusted (P=0.736; OR, 0.94; 95% CI, 0.79–1.13), was demonstrated. Of the 31 observational studies reporting on cardiac arrest rates, 1 before-after study using an aggregated weighted scoring system (Modified Early Warning Score [MEWS]) reported significantly higher cardiac arrest rates in MEWS bands 3 to 4 after intervention, but not in MEWS bands 0 to 2 or 5 to 15, and overall cardiac arrest rate significance was not reported(Subbe 2003, 797); 7 studies with no adjustment demonstrated no significant improvement in cardiac arrest rates after the introduction of a MET system(Kenward 2004, 257; Rothschild 2008, 417; Vazquez 2009, 449; Lim 2011, 373; Shah 2011, 1361; Scherr 2012, 32; Simmes 2012, 20); 15 studies with no adjustment demonstrated significant improvement in cardiac arrest rates after the introduction of a MET system(Bellomo 2003, 283; Dacey 2007, 2076; Offner 2007, 1223; Baxter 2008, 223; Benson 2008, 743; Campello 2009, 3054; Moldenhauer 2009, 164; Konrad 2010, 100; Lighthall 2010, 679; Beitler 2011, R269; Laurens 2011, 707; Moon 2011, 150; Sarani 2011, 415; Rothberg 2012, 98; Al-Qahtani 2013, 506); 4 studies with adjustment demonstrated significant improvement in cardiac arrest rates after the introduction of a MET system both before and after adjustment(Buist 2002, 387; DeVita 2004, 251; Sabahi 2012, 270; Chen 2014, 167); 1 study with contemporaneous controls demonstrated no significant improvement in cardiac arrest rates after the introduction of a MET system both before and after adjustment(Bristow 2000, 236); 1 study with contemporaneous controls demonstrated significant improvement in cardiac arrest rates after the introduction of a MET system both before and after adjustment(Chen 2014, 167); 1 study with adjustment demonstrated significant improvement before adjustment for whole of hospital and non–intensive care unit (ICU) cardiac arrest rates, but only for non-ICU cardiac arrest rates after adjustment(Chan 2008, 2506); and 1 before-after study that presented “after” unadjusted data for cardiac arrest in 3 separate time bands demonstrated significant improvement in time bands 2 and 3.(Santamaria 2010, 445) The heterogeneous nature of the studies prevents pooling of data. However, there is a suggestion of a reduced incidence of cardiac/respiratory arrest in those hospitals that introduce a MET service, and a suggestion of a dose-response effect, with higher-intensity systems (eg, higher MET calling rates, senior medical staff on MET teams) being more effective.
Treatment Recommendation:
We suggest that hospitals consider the introduction of an EWS/response team/MET system to reduce the incidence of IHCA and in-hospital mortality (weak recommendation, low-quality evidence). Values, Preferences, and Task Force Insights This recommendation places a high value on the outcomes—the prevention of IHCA and death—relative to the likely substantial cost of the system. Such a system should provide a system of care that includes (a) staff education about the signs of patient deterioration; (b) appropriate and regular vital signs monitoring of patients; (c) clear guidance (eg, via calling criteria or early warning scores) to assist staff in the early detection of patient deterioration; (d) a clear, uniform system of calling for assistance; and (e) a clinical response to calls for assistance. The best method for the delivery of these components is unclear.(Mancini 2010, S539) The “Recommended Guidelines for Monitoring, Reporting, and Conducting Research on Medical Emergency Team, Outreach, and Rapid Response Systems: An Utstein-Style Scientific Statement”(Peberdy 2007, 2481) should be used by hospitals to collect the most meaningful data to optimize system interventions and improve clinical outcomes.
CoSTR Attachments:
Characteristics_of_included_MET_studies 29_12_14 _n.docx    
Grade Table 25-2_n.docx    
mem_Grade evidence to decision framework _n.docx    

 Contact Us

 
If you have any comments or questions on this page, please email us at: