Background
Methods
Systematic review objectives
Search strategy for identification of studies
Data abstraction
Internal validity and risk of bias assessment
Data analysis
Results
Search results
Author | Source | Intent | Design | Population | (n) | Strain Measure |
---|---|---|---|---|---|---|
Ahmed [1] | Abstract | Quality | Case-control | Adult | 161 | After-hours discharge queuing |
Ahrens [2] | Abstract | Research | Cohort | Pediatric | 764 | ICU census |
Al-Jaghbeer [3] | Full text | Quality | Cohort | Adult | 136 | After-hours discharge ICU readmission |
Amaravadi [4] | Full text | Research | Cohort | Adult | 366 | Nurse-to-patient ratio |
Aytekin [5] | Full text | Research | Correlational | Neonatal | 80 | Burnout |
Azevedo [6] | Abstract | Research | Cohort | Adult | 1329 | ICU acuity ICU readmission |
Beck [7] | Full text | Research | Cohort | Adult | 1654 | After-hours discharge ICU acuity |
Brown [8] | Full text | Research | Cohort | Adult | 268,824 | ICU readmission |
Brown [9] | Full text | Research | Cohort | Adult | 214,692 | ICU readmission |
Chalfin [10] | Full text | Research | Cross-sectional | Adult | 50,322 | Queuing |
Cooper [11] | Full tText | Research | Cohort | Adult | 103,984 | ICU acuity ICU readmission |
Czaja [12] | Full text | Research | Cohort | Pediatric | 111,923 | After-hours discharge ICU readmission |
Dara [13] | Full text | Research | Cohort | Adult | 2492 | ICU acuity Standardized mortality rate |
Duke [14] | Full text | Research | Cohort | Adult | 697 | After-hours discharge |
Duke [15] | Full text | Research | Cross-sectional | Adult | 24,935 | ICU readmission ICU transfer Queuing Surgery cancellation |
Duke [16] | Full text | Research | Cohort | Adult | 3004 | ICU census ICU refusal Surgery cancellation |
Frankel [17] | Full text | Quality | Cohort | Adult | 4956 | ICU readmission |
Frisho-Lima [18] | Full text | Research | Cohort | Adult | 127 | ICU census |
Gajic [19] | Full text | Research | Cohort | Adult | 1131 | ICU acuity ICU readmission |
Gantner [20] | Full text | Research | Cohort | Adult | 109,384 | After-hours discharge ICU readmission |
Goldfrad [21] | Full text | Research | Case-control | Adult | 2269 | After-hours discharge |
Gopal [22] | Abstract | Research | Case-control | Adult | 1257 | After-hours discharge ICU readmission |
Harris [23] | Abstract | Research | Cohort | Adult | 13,086 | Queuing |
Heneghan [6] | Abstract | Research | Cohort | Pediatric | 373 | ICU readmission |
Hung [24] | Full text | Research | Cohort | Adult | 1242 | ICU acuity Queuing |
Iwashyna [25] | Full text | Research | Case-control | Adult | 200,499 | ICU acuity ICU census |
Joynt [26] | Full text | Research | Cohort | Adult | 624 | ICU census Standardized mortality rate |
Kramer [27] | Full text | Research | Cohort | Adult | 369,129 | ICU acuity ICU readmission |
Laupland [28] | Full text | Research | Cohort | Adult | 7380 | After-hours discharge ICU acuity |
Leary [29] | Full text | Quality | Modeling | Adult | 3101 | ICU census Standardized mortality rate Workload |
Lim [30] | Full text | Research | Cohort | Adult | 70 | ICU acuity |
Liu [31] | Full text | Research | Case-control | Adult | 6369 | ICU acuity |
Louriz [32] | Full text | Research | Cohort | Adult | 398 | ICU acuity ICU census |
Nathanson [33] | Full text | Research | Cohort | Adult | 124,855 | ICU acuity |
Parker [34] | Abstract | Research | Cohort | Adult | 255 | Queuing |
Pozzesseres [6] | Abstract | Research | Cohort | Adult | 210 | Queuing |
Priestap [35] | Full text | Research | Cohort | Adult | 47,062 | After-hours discharge ICU acuity |
Pronovost [36] | Full text | Research | Cohort | Adult | 2982 | Daily rounds by intensivist Nurse-to-patient ratio |
Rosenberg [37] | Full text | Research | Cohort | Adult | 4208 | ICU acuity ICU readmission Standardized mortality rate |
Ruse [38] | Abstract | Quality | Cohort | Adult | Unknown | After-hours discharge ICU acuity Queuing |
Santamaria [39] | Full text | Research | Cohort | Adult | 10,221 | After-hours discharge ICU acuity |
Singh [40] | Full text | Research | Cohort | Adult | 2300 | After-hours discharge ICU acuity ICU readmission |
Stelfox [6] | Abstract | Research | Cohort | Adult | 32,234 | ICU readmission |
Tobin [41] | Abstract | Research | Cohort | Adult | 10,903 | After-hours discharge Early ICU discharge ICU acuity Queuing |
Town [42] | Full text | Research | Cohort | Adult | 60,355 | ICU census ICU readmission Turnover Workload |
Tucker [43] | Full text | Research | Cohort | Neonatal | 13,334 | ICU census Nurse-to-patient ratio Turnover |
Wagner [44] | Full text | Quality | Cohort | Adult | 200,730 | ICU acuity ICU census ICU readmission Turnover |
West [45] | Full text | Research | Cross-sectional | Adult | 38,165 | ICU census Nurse-to-patient ratio Turnover Workload |
Yergens [38] | Abstract | Research | Cohort | Adult | 1770 | ICU census |
Amarasigham [46] | Full text | Quality | Quality Improvement | Adult | Unknown | Queuing |
Barado [47] | Abstract | Quality | Modeling | Adult | 6300 | Early ICU discharge ICU refusal |
Study quality
Indicators of capacity strain
Quality indicator | description | Definitions used in the literature for exposure, outcome and analysis |
---|---|---|
ICU census | ICU bed occupancy ICU bed availability | • Total number of patients who spent at least 2 h in the ICU on the calendar day the patient was admitted. • Bed occupancy. • ICU full and not able to admit or discharge any patients. • No available ICU beds |
Queuing | Time delay in patient ICU admission | • Delay in ICU admission. • Delay in ICU admission > 4 h. • Delay in ICU admission > 6 h. • Delay in ICU admission > 8 h. • Mean time from bed request to ICU transfer. |
Nurse-to-patient ratio | Ratio of nurses to patient for a given ICU. | • Ratio of nurses to beds in an ICU. • Ratio > or < 1:2. • Ratio 1:2 vs. 1:2. |
Daily rounds by intensivist | Daily review of patient’s condition and problem list by MRP. | • No definition provided. |
ICU transfer | Transfer of an ICU patient from one ICU to another. | • Inter-hospital transfer of an ICU patient. |
Acuity | Severity of illness of patients in the ICU. | • APACHE II score. • Acute physiology score. • MPM-0 score. |
After-hours discharges | ICU discharge of a patient to the hospital ward outside of regular hours. | • ICU discharge between 1600 and 0800 h. • ICU discharge between 1800 and 0600 h. • ICU discharge between 2000 and 0800 h. • ICU discharge between 2200 and 0700 h. |
Turnover | The number of new admissions to and discharges from an ICU over a given time period. | • Number of new admissions, discharges and transfers. • Number of new admissions per day. • Number of admissions in a given week. |
Workload | Intensity of bedside nurse work required per patient per unit of time. | • Number of new patient admissions and number of patient-care days. • Volume and pressure of work. • TISS score |
Early ICU discharge | Premature ICU discharge. | • Discharged early but would have benefited from longer ICU stay. |
Refusal rate | A measure of the number of patients referred to but not admitted to the ICU. | • Patients who were referred to but not admitted to the ICU. |
ICU readmission | Patients who have been discharged from the ICU and are readmitted within the same hospitalization. | • ICU readmission within 24 h. • ICU readmission within 48 h. • ICU readmission within 72 h. • Unplanned ICU readmission. |
SMR | Ratio between the observed number of deaths in a study population and the number of deaths that would be expected, based on age and sex-specific rates or severity of illness score. | • Not applicable. |
Burnout | Workplace-related psychological stress leading to healthcare providers perception of emotional exhaustion, depersonalization and lack of personal achievement. | • State characterized by physical and/or psychological fatigue, disappointment, underachievement, tiredness and desire to leave work. |
Job satisfaction | Healthcare workers satisfaction with work and workplace environment. | • Nursing self-reports of either being satisfied, unsatisfied or partially satisfied. |
Surgery cancellation | Elective surgeries that is postponed or cancelled due to ICU bed availability. | • Cancellation of surgery due to lack of ICU bed. • Surgery cancelled or rescheduled. |
National quality measurement and reporting criteria
Categorization of strain measure by the Donabedian framework† | Importance¶ (n = 95) | Scientific acceptability¶ (n = 23) | Usability and feasibility¶ (n = 11) | |||
---|---|---|---|---|---|---|
Quality (n = 42) | Patient-centered outcomes (n = 47) | Healthcare costs (n = 6) | Operational (n = 8) | Integrate into EHR (n = 3) | ||
Structure (n = 30) | ||||||
1. ICU census (n = 13) | 7 | 7 | 1 | 4 | 3 | 1 |
2. Queuing (n = 11) | 2 | 3 | 1 | 3 | 1 | – |
3. Nurse to patient ratio (n = 5) | 2 | 3 | 1 | 2 | 1 | – |
4. Daily rounds by intensivist (n = 1) | 1 | 1 | – | – | – | – |
Process (n = 50) | ||||||
5. ICU transfer (n = 1) | – | – | – | – | – | – |
6. ICU acuity (n = 21) | 10 | 10 | – | 2 | – | – |
7. After-hours discharge (n = 15) | 11 | 11 | 2 | 8 | 1 | – |
8. Turnover (n = 4) | 2 | 2 | – | – | – | – |
9. Workload (n = 4) | 1 | 1 | – | – | 1 | 1 |
10. Early ICU discharge (n = 3) | – | – | – | – | – | – |
11. Refusal rate (n = 2) | – | – | – | – | – | – |
Outcome (n = 30) | ||||||
12. ICU readmission (n = 18) | 7 | 7 | 1 | 2 | – | – |
13. SMR (n = 4) | 1 | 1 | – | – | 1 | 1 |
14. Burnout (n = 2) | – | – | – | 1 | – | – |
15. Job satisfaction (n = 2) | – | – | – | 1 | – | – |
16. Surgery cancellation (n = 2) | – | – | – | – | – | – |
Discussion
Summary of key findings
Context with prior literature
Limitations
Implications for healthcare professionals, health policy and research
Short-term measures (e.g., daily or weekly) | Intermediate term measures (e.g., monthly or quarterly) |
---|---|
ICU acuity | ICU readmission |
After-hours discharges | Burnout |
ICU census | Workplace satisfaction |
Sedation interruption* | Early ICU discharge |
Queuing | Surgery cancellation |
Mobilization* | ICU transfer |
Nurse-to-patient ratio | Refusal rate |
Turnover | Adverse events* |
Mechanical ventilation weaning* | SMR |
Workload | Family satisfaction* |
Daily rounds by intensivist |
Quality Indicator | Proposed aggregate definition | Justification | Proposed benchmark or measure |
---|---|---|---|
ICU census | ICU not able to admit any new patients. | The exact percentage bed occupancy is less important than having capacity to admit. | < 10% of time |
Queuing | Delay in time from orders to admit to ICU to ICU arrival. | Increasing delays for ICU admission result in suboptimal care for these critically ill patients. The most common timeframe in the literature was within 4 h of decision to admit. | < 4 h |
Nurse-to-patient ratio | The number of nurses caring per patient. | A lower ratio of nurses per patient means less time can be spent per individual patient and increases nursing workload. To most common ratio studied in the literature was 1:2. | Adjusted nursing workload of < 1:2 |
Daily rounds by intensivist | Daily bedside visit by MRP to review patients’ medical condition and problem list. | Daily in-person rounds are critical when caring for ICU patients. These should occur daily in a formal fashion. | 100% |
ICU transfer | Inter-hospital transfer of an ICU patient due to lack of capacity. | This definition interplays with that of ICU census; however, it is an extension of the above, indicating that there are no mechanisms for increasing capacity at the strained institution. | None |
ICU acuity | The average severity of illness of patients in the ICU. | More acutely ill patients provide both a physical and mental strain on ICU staff. The APACHE II score was most commonly used in the literature. However, institutionally specific scores may be used as well. | APACHE II Score |
After-hours discharges | Unplanned discharges from the ICU outside of regular hours (as defined in per each individual institution) | Patients discharged outside of regular hours may not be evaluated by medical staff in a timely fashion. There were many definitions of ‘after-hours’ in the medical literature. ‘After-hours’ should relate to individual institutional practices. | None |
Turnover | The number of admissions and discharges from an ICU per 24-h period. | Typically highest patient workload occurs on ICU admission and discharge. | n/a |
Workload | The volume and pressure of work. | Higher workload can lead to increased stress and concerns regarding patient safety. An objective measure of workload is necessary to quantify this variable. | TISS-28 Score |
Early ICU Discharge | Discharge from an ICU earlier than preferable as per the MRP. | Physicians must triage patients at time of ICU capacity strain to ensure that the sickest patients be those located in the ICU. This may require immediate decision-making regarding discharging of less acutely ill patients. | None |
Refusal Rate | The ratio of patients refused entry to the ICU vs. total number of ICU consults. | As strain in the ICU increases, physicians are less likely to admit patients who may not truly require ICU level care. This needs to be balanced with referred patients who do not require ICU level care. | 0% of appropriate ICU consultations |
ICU Readmission | Avoidable ICU readmission within 48 h of discharge as adjudicated by admitting physician. | Most ICU readmissions are unavoidable and hence are not a reflection of ICU strain or quality. However, if an ICU is under strain and patients are discharged prematurely and this results in ICU readmissions, this may be a marker of strain. Avoidable readmissions should be adjudicated as per the admitting physician. | None |
Standardized mortality ratio | Ratio between the observed number of deaths in a study population and the number of deaths that would be expected, based on age and sex-specific rates in a standard population and the age and sex distribution of the study population. | An increasing varying SMR may be related to varying ICU strain. Benchmark is based on data from all Alberta provincial ICUs. | < 15% |
Burnout | Work-related stress leading to feelings of pressured, overwhelmed and desire to leave work. | As workload and patient acuity increases, healthcare providers may themselves feel overwhelmed and unable to carry on work. An objective measure of burnout syndrome (BoS) is necessary to quantify this variable and the Maslach Burnout Inventory has been extensively studied in the literature and may be referenced across ICUs. | Maslach Burnout Inventory |
Job satisfaction | Healthcare workers reporting lack of satisfaction with their job. | With increasing strain and stress at the workplace, there is decreasing satisfaction on the job. An objective measure is necessary to quantify this variable. | Measurement of Job Satisfaction |
Surgery cancellation | Surgeries that require cancelation of rescheduling due to ICU constrains. | Certain elective surgeries necessitate post-operative ICU monitoring. However, in cases of strain, these surgeries may be cancelled or rebooked. | None |
Conclusions
Key messages
-
Strained ICU capacity is associated with alterations in care processes and adverse outcomes.
-
This systematic review has identified and characterized 16 potential indicators of strained ICU capacity.
-
Indicators were variable in their operational definitions and few were evaluated for scientific acceptability, usability or feasibility.
-
The most common indicators of strain showed overlap with recommended ICU key performance indicators (i.e., ICU acuity, ICU readmission, after-hours discharge, and occupancy).
-
Several indicators of strain could readily be implemented and would likely add value, particularly if clustered as a dashboard or index, to provide holistic ICU-specific information on key contributors to strain.