Factors associated with 30-day mortality in medical emergency team patients – a retrospective cohort study

Background: Medical emergency teams (METs) have been implemented to respond more eciently to inpatient crises in order to reduce hospital mortality by the early recognition and treatment of potentially life-threatening conditions. The objective of this study was to identify factors associated with 30-day mortality risk among patients assessed by the MET. Methods: Observational retrospective register study in a tertiary university hospital in Sweden, comprising 2,601 patients, assessed by the MET from 2010 to 2015. Patient registry data at the time of MET assessment was analysed from an epidemiological perspective, using univariable and multivariable analyses with death within 30 days as the endpoint. Results: The overall 30-day mortality was high (29.0%) and almost twice as high on medical wards as on surgical wards (37.1% vs 19.8%). The acute medical conditions associated with the highest 30-day mortality were gastroenteritis (50.0%), acute coronary syndrome (43.0%), cardiac failure (41.1%) and renal failure (40.4%). The multivariable analysis indicated that factors with increased 30-day mortality were reected in age, type of ward, vital parameters, laboratory ndings, previous medical history and acute medical condition. Apart from age, the factor associated with the highest odds ratio for 30-day mortality was hypoglycaemia (OR 4.30 (95% CI 2.10, 8.81)). Conclusions: We found that factors reecting the patients' age, co-morbidity and acute medical condition, along with clinical ndings such as vital parameters and laboratory abnormalities, and also the type of ward for admittance, all contributed to identifying patients at risk of death within 30 days after MET assessment. Overall, medical ward patients have higher 30-day mortality than surgical ward patients. in vital parameters reecting respiratory appeared be more alarming than waning circulatory


Background
Over the years, healthcare has evolved rapidly with increasingly older and more ill patients being attended to, resulting in enormous demand. Rapid response systems (RRS), including medical emergency teams (MET), provide frail patients in general wards with continuous access to critical care expertise when early signs of clinical deterioration are recognised (1,2). Despite this, intensive care resources are insu cient (3,4). It would, therefore, be desirable to re ne the selection of patients who would bene t most from admission to the intensive care unit (ICU).
The objective of this study was to identify factors associated with 30-day mortality risk among patients assessed by the MET. Patient registry data at the time of MET assessment was analysed from an epidemiological perspective, using univariable and multivariable analyses with death within 30 days as the endpoint. We believe that early identi cation of patients with high mortality risk can be valuable in the guidance of further therapeutic efforts and adaptation to available resources. We hypothesised that it is possible to identify factors at the time of MET assessment associated with an increased mortality risk during the subsequent 30 days. Potential risk factors could then be considered more systematically as decision support when prioritising and optimising the early chain of care.

Settings
The study was performed at Sahlgrenska University Hospital in Gothenburg, Sweden. The hospital provides specialised care and is the trauma referral centre for the region. It has some 700 beds available for close to 50,000 admissions and 18,000 surgical procedures each year. The MET service was introduced at Sahlgrenska in 2005. Since 2007, the MET service has operated at full scale on all nursing wards, except for thoracic surgery wards, receiving approximately 600 consultation calls annually. A breakdown of the number of MET assessments versus hospital admissions is reported (Additional le 1).

MET system
The MET system is activated by ward staff when patients develop prede ned alterations in vital parameters, or when any of the staff feels worried about the patient (5). The MET system is designed to be activated in patients with abnormal vital parameters. In the event of an immediate life-threatening condition, the cardiac arrest team (CAT) should be alerted.
During the study period, the MET service was available 24 hours/day, all week. MET included an intensive care specialist at consultant level during the day and an intensive care trainee at night, plus an intensive care nurse. The CAT service operated similarly but consisted of three doctors trained in intensive care, internal medicine and cardiology, respectively and usually an anaesthetic nurse. The MET system utilised a single-parameter track-and-trigger system, with the following activation criteria: Respiratory rate <8 or >30 breaths/minute Systolic blood pressure <90 mmHg Heart rate <40 or >130 beats/minute Decreased level of consciousness Serious concern regarding the patient's health Until September 2013,' threatened airway' was included as a criterion. However, due to the seriousness of this condition, it was removed to be handled by the CAT instead. During the time of the study, individual, need-based ordinations for checking vital parameters were applied. Given the immense diversity of ward patients, a universal praxis was refrained from. Depending on clinical ndings, ward nurses were delegated to administer oxygen to patients with hypoxia.

Study design
The study was conducted as a retrospective, observational study of registry data on MET assessed patients, 1 January 2010 to 31 December 2015. The outcome was 30-day mortality.

Study population
To be included, patients had to be assessed by the MET on general wards and registered in a standardised protocol. The patient had to be 18 years or older, with known 30-day survival status. In the event of repeated MET assessments, only the rst MET assessment during each hospital episode was included in the main study ( Figure 1).
To investigate whether patients requiring additional MET assessments during the same admission period differed from patients requiring only one MET assessment. These two patient groups were compared regarding age, gender and previous medical history (Additional le 2).

Data collection
Data were collected from the assessment protocol, supplemented with electronic medical records.
Baseline characteristics, type of ward, previous medical history, the reason for MET call, vital parameters at MET arrival, laboratory markers from up to 48 hours before and 6 hours after MET activation, acute medical condition, limitation of medical therapy (LOMT), potential ICU admission and primary diagnosis were recorded. Thirty-day survival status was obtained and con rmed from the Swedish population registry (Additional le 3).

Statistical analysis
Logistic regression was used to calculate age-adjusted p-values for the association of each variable with 30-day mortality. Due to the amount of missing data for several of the variables, multiple imputations were used for multivariable analysis. Missing data were assumed to be missing at random (MAR) and 50 imputed datasets were generated using the Markov Chain Monte Carlo (MCMC) method using the expectation-maximisation (EM) algorithm. Rubin's rules were used to pool results from the imputed datasets. To exclude the possibility that missing data pattern was missing completely at random (MCAR), we compared cases with no missing data with incomplete cases and found several signi cant differences. The assumption of a MAR pattern was shown to be valid by examining the associations between the missingness of each variable with other variables. To identify independent factors associated with 30-day mortality, we started with a model including age and all variables in Tables 1-5 with an age-adjusted p <0.20. Collinearity was checked by association measures between variables and by inspecting the variance in ation factor, condition index and eigenvector proportions in a multiple linear regression model, including all these candidate variables. Thresholds for collinearity were a variance in ation factor above 5 and/or condition index above 10. Multiple logistic regression was performed in each of the 50 imputed datasets and the variable with the highest p-value in the pooled result was excluded from the model. A new regression analysis was then performed in each imputed dataset. Of the remaining variables, the one with the highest p-value in the pooled result was excluded. This procedure was repeated until all the remaining variables yielded a p-value below 0.01 in the pooled result. As a sensitivity analysis to see whether our ndings were consistent without imputation, we also performed a multiple logistic regression analysis with a backward stepwise selection of variables in a similar manner and including only complete cases. Two-sided tests were used, and p-values of <0.01 were considered statistically signi cant. Analyses were performed using SAS version 9.4 for Windows software (SAS Institute, Cary, NC, USA).

Study participants and outcome
A total number of 2,601 patients ful lled inclusion criteria ( Figure 1). The study population comprised patients 18 to 99 years of age (mean=65.7, SD 16.8), of which 44.3% were female. Fewer than half the patients (42.5%) were transferred to the ICU. Overall 30-day mortality was 29.0%. Patients with palliative decisions and LOMT demonstrated a signi cantly higher 30-day mortality (65.5%), in comparison to patients without any treatment restrictions (21.2%). There were, however, no signi cant differences in 30day mortality with regards to gender or level of care (age-adjusted p=0.37 and 0.31, respectively) (Additional le 4).

Type of ward
Most of the patients assessed by the MET were admitted to general medicine wards, followed by general surgery and neurological wards. The highest 30-day mortality was found in geriatric wards, followed by respiratory medicine and oncology wards. Overall, there was a similar number of patients from any medical and any surgical wards. MET assessed patients on surgical wards had signi cantly lower 30-day mortality (Table 1).
Results presented as number (percent) * Neurological wards include neurology, spinal injury and neurosurgery patients # Age-adjusted p-value for association with 30-day mortality

Previous medical history
The most frequently reported conditions in the patients' previous history were hypertension, cancer and lung diseases. Previous conditions associated with the highest 30-day mortality were cardiac failure, followed by haematological disease, angina pectoris and pulmonary disease. The following previous conditions were signi cantly associated with increased age-adjusted mortality during the subsequent 30 days: cancer, haematological disease, pulmonary disease, and liver disease. (Table 2).
Results presented as number (percent) * Information on previous medical history was missing for one patient who died within 30 days ** Including; cardiac arrhythmias, valvular heart diseases, pericardial disorders, cardiogenetic disorders or congenital heart defects, among others # Age-adjusted p-value for association with 30-day mortality Acute medical condition The acute conditions most frequently associated with MET activation were sepsis and pneumonia. Acute conditions associated with the highest 30-day mortality were gastroenteritis, acute coronary syndrome, cardiac failure and renal failure. The following acute medical conditions were associated with increased age-adjusted mortality during the subsequent 30 days: cardiac failure, pneumonia and renal failure. Other infections (exempli ed in the footnote), postoperative infection and other postoperative complications (exempli ed in the footnote) were associated with a decreased mortality (Table 3).    The most frequent laboratory alteration was hyperglycaemia, followed by low haemoglobin, hypoxaemia and elevated serum creatinine. Laboratory ndings associated with the highest 30-day mortality were hypoglycaemia, hypernatraemia, hyperkalaemia, acidosis and hyperlactatemia. The same laboratory ndings were also associated with signi cantly increased age-adjusted mortality, along with elevated serum creatinine and hypoxaemia (Table 4).   The most frequent abnormalities in terms of vital parameters were hypoxia and tachypnoea. Patients who presented with these ndings also had the highest 30-day mortality. The following vital parameters were associated with a signi cantly higher age-adjusted mortality risk during the subsequent 30 days: hypoxia, tachypnoea, tachycardia and unconsciousness (Table 5).    * Number of patients for whom information was missing in the two groups, respectively # Age-adjusted p-value for association with 30-day mortality POX, pulse oximetry; RR, respiratory rate; HR, heart rate; SBP, systolic blood pressure; RLS, reaction level scale Independent factors associated with mortality No important collinearity was found among candidate variables for association with 30-day mortality in the multivariable analysis. When using multiple imputations, age, type of ward, vital parameters, laboratory ndings, previous medical history, and acute medical condition all contributed to the prediction of death. Apart from age, the factors with the highest odds ratio for death within 30 days were hypoglycaemia, haematological disease and hypoxia ( Table 6).    In order to illustrate in more detail the time of death after MET assessment, cumulative mortality curves for the thirteen factors independently associated with 30-day mortality ( Table 6) are presented (Additional le 5). Overall, approximately half of the deaths occurred within the rst four days after MET assessment (Figure 2).

Discussion
The three main ndings in this study were that (1) overall mortality among patients triggering the RRS was high. More than one out of four patients died within 30 days, and the higher the age, the greater the mortality, independent of gender. (2) The previous and acute medical conditions of patients, such as haematological disease, liver disease, cancer and renal failure, were independently and signi cantly associated with increased 30-day mortality. (3) The laboratory ndings corresponding to the highest mortality risk were hypoglycaemia, hyperlactatemia, hypernatremia and hyperkalemia. In addition, this study implied the importance of several other factors associated with mortality in clinically deteriorating patients, including abnormal vital parameters such as hypoxia and tachypnoea, level of consciousness and type of ward activating the RRS.
The noticeably high mortality rate in this MET population could be explained by the possibly lifethreatening situation of clinical deterioration in combination with advanced age and numerous comorbidities. As stated in previous studies, MET patients are in the poorest condition among hospitalised patients with high in-hospital and 30-day death rates (1,6). Analogously, the type of ward demonstrating by far the highest risk of death among admitted patients was the geriatric wards. More than half of the clinically deteriorated geriatric patients triggering the MET died within 30 days after assessment. Given the indisputable importance of age in relation to survival, the potentially bene cial contribution of the patient's age as a trigger component in the early warning system cannot be ignored.
The association between type of ward and mortality re ects the commonly found medical conditions. It appears that medical and surgical wards tend to utilise the RRS to about the same extent, although the difference in outcome was striking in our study. The overall 30-day mortality on medical wards was almost twice as high as that on surgical wards. Consequently, the patient's place of care may be indicative of the end of the course, in terms of mortality, as medical ward patients proved to have a signi cantly increased risk of death. Thus, the opposite proved to be true for surgical ward patients. The difference in 30-day mortality between surgical and medical ward patients could be explained in part by the fact that surgical patients tended to have less co-morbidity and more of an isolated problem (Additional le 6).
The difference in outcome regarding the patient's place of care is further re ected when analysing the patient's previous and acute medical conditions with 30-day mortality. The most aggravating previous history was primarily internal medicine diseases, such as cardiac failure, haematological disease, angina pectoris and pulmonary disease. Correspondingly, the acute medical conditions associated with the highest 30-day mortality were predominantly internal medicine diseases, such as gastroenteritis, acute coronary syndrome, cardiac failure and renal failure. Surgically related conditions, such as postoperative infection and other postoperative complications, were signi cantly associated with lower mortality risk.
The common denominator among the surgical patients appeared to be a transient or reversible illness with a good prognosis when treated correctly.
The most frequently used trigger criterion was POX <90% (Additional le 7). Interestingly, the incidence of POX <90% decreased by almost 10 per cent units between MET activation and arrival (Table 5), which is believed to depend on the general delegation of oxygen administration issued to the ward nurses and possibly also oxygen treatment recommendations over the phone pending the arrival of the MET. Despite this initial sign of improved optimisation, hypoxia and tachypnoea were associated with the highest 30day mortality among vital parameters, which strengthens the ndings in previous studies (7). Curiously, circulatory parameters did not play as important a role in predicting outcome in the MET patient population.
Caution should be taken concerning the fact that one of the vital parameters of the utmost importance in predicting 30-day mortality, i.e. the respiratory rate, was the parameter least often measured by the METmissed in more than every fth patient. Considering the signi cant mortality risk when a patient presents with tachypnoea, we call for more attention to be paid to monitoring this vital parameter (8)(9)(10).
Despite hypoglycaemia being the most rarely encountered abnormal laboratory nding, it was associated with the highest mortality risk of all measured risk factors. It is worth noting that numerous studies have shown that hyperglycaemia is strongly associated with increased morbidity and mortality, e.g. in acute coronary syndrome, in admitted patients, regardless of whether or not they were diagnosed with diabetes mellitus (11,12). Nonetheless, our nding is also well-founded in the literature, where hypoglycaemia has been shown in several studies to be an independent risk factor for death in patients with acute illness (13)(14)(15). Hence, it appears that disturbances in glucose metabolism, regardless of direction, signify an increased risk of adverse outcomes among critically ill patients. Other laboratory ndings associated with poor outcome were hyperlactatemia, hypernatraemia, hyperkalaemia, acidosis, hypoxaemia and elevated creatinine. Apart from creatinine, all the mortality-indicative laboratory values are found in regular blood gas analyses. Our data, however, revealed that arterial blood gases were missing during the care event for 40% of the patients, which leads us to speculate about whether more frequent blood gas sampling in clinically deteriorating patients would be bene cial for the early detection of severe illness and, by extension, improved outcome.
In order to demonstrate independent associations of epidemiological factors in the MET patient population, multivariable analyses were performed. Due to the large quantity of included variables, more than half the study population was excluded from the complete data analysis, as a result of missing data on one or more of these variables (Additional le 8). Therefore, a multivariable analysis using multiple imputations was performed ( Table 6). Regardless of the method, the results of both multivariable analyses were consistent with those of the analyses of each variable separately. As previously indicated, hypoglycaemia proved to be a major risk factor for death within 30 days, followed by, among several others, haematological disease and hypoxia. Again, attendance on surgical wards was a favourable factor of reduced mortality.

Clinical implications
Medical ward patients have a higher risk of death within 30 days than surgical ward patients.
Abnormalities in vital parameters re ecting respiratory distress appear to be more alarming than waning circulatory functions. Altogether, the elderly MET patients, patients admitted to medical wards and patients presenting with abnormal respiratory function or renal failure, and patients with a history of haematological disease, liver disease or cancer have the highest risk of death within 30-days. For this reason, reducing the threshold for intensive care among these patient groups may be bene cial from a survival point of view, provided that no LOMT are applied. Factors associated with mortality emanate from many diverse sources. Using a standardised risk score system at MET assessment as a basis for estimation of the mortality risk should, therefore, be favourable from a prioritisation and optimisation perspective.

Study strengths and limitations
Strengths: the study was population-based with well-de ned inclusion and exclusion criteria and a relatively large sample size. Furthermore, it was consecutive, and all cases were evaluated for inclusion. Moreover, it was chart based, with all cases handled manually.
Limitations: data were limited to 2010-2015, and new conditions may have emerged since then. The 'MET dose', calculated as the number of MET assessments divided by the number of hospital admissions (approximately 12/1,000), was low in comparison to studies in other healthcare systems, possibly indicating an ine ciency in the system (16). Despite this, our 'MET dose' is higher than the dose previously reported in a before-and-after trial in Sweden, where the implementation of MET was associated with a signi cant improvement in cardiac arrest rate and in-hospital mortality (17). Given the retrospective design, it was not possible to check and correct for afferent limb failure, i.e. delayed activation of the MET (18). Further, it was a single-centre study using a single-parameter system, not fully transferable to hospitals with different routines. Since it was a retrospective study, we were not able to control for unmeasured factors. Also, for several of the variables, the number of missing data was substantial, although we tried to handle this by using multiple imputation methods in the multivariable analysis. In a retrospective register study, it is not possible to draw any conclusion regarding the cause and the effect. We are only able to describe associations.

Conclusions
The MET population remains an exposed category amongst hospitalised patients. The overall 30-day mortality was 29.0%. We found that factors re ecting the patients' age, co-morbidity and acute medical condition, along with clinical ndings such as abnormal vital parameters and laboratory abnormalities, and also the type of ward for admittance, all contributed to identifying patients at risk of death within 30 days after MET assessment.

Declarations Funding
The study was supported by grants from the Swedish state under the agreement between the Swedish government and the county councils (the ALF agreement: ALF GBG-716901, ALF GBG-74200), the Swedish Heart and Lung Foundation and the Gothenburg Medical Society.