Systematic review and evaluation of physiological track and trigger warning systemsfor identifying at risk patients on the ward
Electronic Supplementary Material (ESM)
Search strategy for electronic databases
Ovid MEDLINE(R) <1966 to November Week 3 2004>
- *health status indicators/ or exp *"severity of illness index"/ or *sickness impact profile/ (8850 records)
- *Risk Assessment/ (5692)
- severity of illness ind$.tw. (90)
- health status ind$.tw. (514)
- risk assess$.tw. (11356)
- sickness impact profile$.tw. (753)
- early warning score$.tw. (9)
- ews.tw. (648)
- (track and trigger).tw. (21)
- modified early warning score$.tw. (7)
- mews.tw. (9)
- *Point-of-Care Systems/ (1295)
- point-of-care system$.tw. (38)
- point of care system$.tw. (38)
- ((trigger or calling) adj5 criteria).tw. (39)
- serious$ ill$.tw. (2790)
- or/1-16 (30375)
- exp *Critical Care/ (15634)
- critical care.tw. (7683)
- intensive care.tw. (41157)
- exp *Intensive Care Units/ (14487)
- (intensive care adj2 (unit$ or ward$ or clinic$ or department$ or section$)).tw. (29240)
- exp *Emergency Service, Hospital/ (14197)
- hospital emergency service$.tw. (155)
- medical emergency team$.tw. (36)
- met.tw. (44331)
- hospital emergency team$.tw. (2)
- patient emergency team$.tw. (0)
- exp *Patient Care Team/ (13621)
- patient care team$.tw. (135)
- patient at risk team$.tw. (2)
- patient-at-risk team$.tw. (2)
- par.tw. (3491)
- patient-at-risk.tw. (1703)
- (outreach adj (service$ or team$)).tw. (256)
- shock team$.tw. (2)
- or/18-36 (134622)
- 17 and 37 (1707)
- limit 38 to (human and "all adult (19 plus years)" and yr=1990-2005) (691)
- from 39 keep 1-691 (691)
Ovid MEDLINE(R) In-Process & Other Non-Indexed Citations <January 14, 2005>
- health status ind$.tw. (17)
- severity of illness ind$.tw. (3)
- risk assess$.tw. (668)
- sickness impact profile$.tw. (13)
- early warning score$.tw. (0)
- ews.tw. (75)
- (track and trigger).tw. (2)
- modified early warning score$.tw. (0)
- mews.tw. (1)
- point of care system$.tw. (5)
- point-of-care system$.tw. (5)
- serious$ ill$.tw. (63)
- ((trigger or calling) adj5 criteria).tw. (1)
- or/1-13 (848)
- critical care.tw. (286)
- intensive care.tw. (1587)
- (intensive care adj2 (unit$ or ward$ or clinic$ or department$ or section$)).tw. (1272)
- hospital emergency service$.tw. (4)
- medical emergency team$.tw. (5)
- met.tw. (2114)
- hospital emergency team$.tw. (1)
- patient emergency team$.tw. (0)
- patient care team$.tw. (5)
- patient at risk team$.tw. (0)
- patient-at-risk team$.tw. (0)
- par.tw. (342)
- (outreach adj (service$ or team$)).tw. (15)
- shock team$.tw. (0)
- patient at risk$.tw. (69)
- patient-at-risk$.tw. (69)
- or/15-30 (4312)
- 14 and 31 (20)
- from 32 keep 1-20 (20)
Embase (NHS Dialog)
- Disease Severity.MJ
- Injury Scale.MJ
- Heath Survey.MJ
- Health Status.DE
- or/1-4
- ind$
- 5 and 6
- sickness impact profile$
- severity of illness (indices or index$)
- health status (indices or index$)
- early warning score$
- modified early warning score$
- track and trigger
- point of care system$
- point-of-care system$
- (trigger OR calling) criteria
- or/8-16
- 7 or 17
- Intensive Care.DE
- Intensive Care Unit.DE
- Emergency Health Service.DE
- critical care OR intensive care
- hospital emergency service$
- medical emergency service$
- hospital emergency team$
- patient emergency team$
- patient at risk team$
- patient-at-risk team$
- outreach (service$ OR team$)
- shock team$
- or/19-30
- 18 and 31
CINAHL - Cumulative Index to Nursing & Allied Health Literature <1982 to December Week 2 2004>
- *Health Status Indicators/ (558)
- *"Severity of Illness Indices"/ (218)
- *Sickness Impact Profile/ (16)
- *Risk Assessment/ (1616)
- severity of illness ind$.tw. (51)
- health status ind$.tw. (111)
- risk assess$.tw. (1424)
- sickness impact profile$.tw. (593)
- early warning score$.tw. (1)
- ews.tw. (10)
- (track and trigger).tw. (6)
- modified early warning score$.tw. (1)
- mews.tw. (6)
- point of care system$.tw. (15)
- point-of-care system$.tw. (15)
- serious$ ill$.tw. (492)
- ((trigger or calling) adj5 criteria).tw. (7)
- or/1-17 (4566)
- *Critical Care/ (2184)
- critical care.tw. (5564)
- intensive care.tw. (6769)
- *Intensive Care Units/ (1528)
- (intensive care adj2 (unit$ or ward$ or clinic$ or department$ or section$)).tw. (5075)
- exp *Emergency Service/ (4138)
- hospital emergency service$.tw. (12)
- medical emergency team$.tw. (9)
- met.tw. (5084)
- hospital emergency team$.tw. (0)
- patient emergency team$.tw. (0)
- *Multidisciplinary Care Team/ (2428)
- patient care team$.tw. (44)
- patient at risk team$.tw. (1)
- patient-at-risk team$.tw. (1)
- par.tw. (290)
- patient-at-risk.tw. (271)
- patient at risk.tw. (271)
- (outreach adj (service$ or team$)).tw. (149)
- shock team$.tw. (0)
- or/19-38 (25007)
- 18 and 39 (244)
- limit 40 to ((adult <19 to 44 years> or middle age <45 to 64 years> or aged <65 to 79 years> or "aged <80 and over>") and yr=1990 - 2005) (98)
- from 41 keep 1-98 (98)
PsychInfo
- health status ind* (116)
- severity of illness ind* (24)
- sickness impact profile* (196)
- risk assess* (1849)
- early warning score* (0)
- ews (17)
- track and trigger (2)
- modified early warning score* (0)
- mews (11)
- point of care system* (4)
- point-of-care system* (0)
- (trigger or calling) near criteria (52)
- critical care (411)
- intensive care (1110)
- intensive near (unit* or ward* or clinic* or department* or section*) (1808)
- hospital emergency service* (37)
- medical emergency team* (3)
- hospital emergency team* (0)
- patient emergency team* (0)
- patient care team* (11)
- patient at risk team* (0)
- patient-at-risk team* (0)
- patient at risk (41)
- patient-at-risk (0)
- outreach near (service* or team*) (699)
- shock team* (1)
- #1 or #2 or #3 or #4 or #5 or #6 or #7 or #8 or #9 or #11 or #12 (2266)
- #13 or #14 or #15 or #16 or #17 or #18 or #19 or #20 or #21 or #22 or #23 or #24 or #25 or #26 (3179)
- #27 and #28 (27)
(Note: “#1 or #2 or #3” indicates that all the results from searches 1 to 3 should be returned and is equivalent to the MEDLINE syntax “or/1-3”)
Cochrane Library (Wiley InterScience website)
- Health Status Indicators (MeSH)
- APACHE (MeSH)
- Severity of Illness Index (MeSH)
- Sickness Impact Profile (MeSH)
- Karnofsky Performance Status (MeSH)
- or/1-5
- Emergency Service, Hospital (exp MeSH)
- Point-of-Care Systems (exp MeSH)
- Critical Care (exp MeSH)
- medical emergency team*
- hospital emergency team*
- patient at risk team*
- patient-at-risk team*
- outreach team*
- outreach service*
- or/7-15
- 6 and 16
Web of Science
- severity of illness ind*
- health status ind*
- risk assess*
- sickness impact profile*
- early warning score*
- ews
- modified early warning score*
- mews
- track and trigger
- point of care system*
- point-of-care system*
- (trigger or calling) SAME criteria
- or/1-13
- critical care
- intensive care
- hospital emergency service*
- medical emergency team*
- hospital emergency team*
- patient emergency team*
- patient care team*
- patient at risk team*
- patient-at-risk team*
- outreach SAME (service* or team*)
- or/14-23
- 13 and 24
List of professional bodies and experts contacted toreview the completeness ofthe list of identified papers
- Intensive Care Society
- RCN Critical Care Forum
- RoyalCollege of Anaesthetists
- British Association of Critical Care Nurses
- National Outreach Forum
- RoyalCollege of Physicians
- RoyalCollege of Physicians and Surgeons of Glasgow
- RoyalCollege of Surgeons of England
- Steering Group members/research team (n = 21)
- Other clinical experts (n = 9)
Table S1: Modified DoCDat criteria for assessing the coverage (A–D) and accuracy (E–F) of physiological track and trigger warning system databases
Level 1 / Level 2 / Level 3 / Level 4A. Was the spectrum of patients representative of the patients who will be monitored with the TT in practice? / No evidence or unlikely to be representative.
(e.g. only patients referred to CCOS) / Some evidence that eligible population is representative.
(e.g. all patients seen by CCOS with CC follow-up can be identified) / Good evidence that eligible population is representative.
(e.g. all patients on selected wards) / Total population in your current setting.
(e.g. all patients on all wards that could be attended by CCOS)
B. Were selection criteria clearly defined? / No / Yes
C. Completeness of TT variables / Only summary TT variables or scores. / Summary TT variables or scores;
at least admission to CC and death were recorded as minimum outcomes / Raw physiological data;
at least admission to CC and death were recorded. / Raw physiological data;
all outcomes;
important confounders
D. Completeness of data (% variables at least 95% complete) / Few (<50%) / Some (50-79%) / Most (80-97%) / All or almost all (>97%)
E. Use of explicit definitions and rules for variables / None / Some (<50%) / Most (50 -97%) / All or almost all (>97%)
F. Extent to which data are validated / No validation / Range or consistency checks / Range and consistency checks / Range and consistency checks plus external validation using alternative source
CC: critical care; CCOS: critical care outreach service; TT: physiological track and trigger warning system.
Table S2: Sensitivity and specificity of physiological track and trigger warning systems by hospital and patient subgroup
Hospital / Sub-groups / Patientsn (%) / Sensitivity
(95% CI) / Specificity
(95% CI) / PPV
(95% CI) / NPV
(95% CI) / Prevalence
(95% CI)
A / CC follow-up / 701 (74.1) / 32.8
(21.3,46.0) / 85.5
(82.5,88.1) / 17.7
(11.2,26.0) / 93.0
(90.7,95.0) / 8.7
(6.7,11.0)
Referral / 245 (25.9) / 48.2
(41.8, 54.6)
B / CC follow-up / 209 (44.4) / 42.9
(17.7,71.1) / 95.4
(91.2,98.0) / 42.9
(17.7,71.1) / 95.4
(91.2,98.0) / 7.4
(4.1,12.1)
Referral / 262 (55.6) / 31.8
(26.4,37.6)
C / Referral / 405 (100) / 26.4
(22.2, 31.0)
D / CC follow-up / 1098 (46.3) / 3.2
(0.1,16.7) / 99.2
(98.4,99.6) / 10.0
(0.3,44.5) / 97.2
(96.1,98.1) / 2.8
(1.9,4.0)
Referral / 1273 (54.7) / 35.1
(32.5, 37.8)
E / CC follow-up / 1119 (34.3) / 43.3
(32.9,54.2) / 90.4
(89.0,91.7) / 17.4
(12.7,23.0) / 97.2
(96.3,97.9) / 4.5
(3.6,5.4)
Referral / 2149 (65.7) / 43.2
(40.0,46.3)
F / CC follow-up / 289 (87.6) / 60.9
(48.4,72.4) / 64.2
(57.5,70.6) / 35
(26.5,44.2) / 83.8
(77.4,89.1) / 24
(19.2,29.4)
Referral / 41 (12.4) / 29.3
(16.1,45.5)
G / MAU patients / 750 (100) / 65.1
(49.1,79.0) / 65.8
(62.1,69.3) / 10.4
(7.0, 4.6) / 96.9
(94.9,98.2) / 5.7
(4.2,7.6)
H / All CCOS / 1051 (100) / 36.7
(33.6,39.8)
I / Referral* / 2463 (100) / 64.7
(62.8,66.6)
J / CC follow-up / 1512 (78.4) / 15.8
(9.8,23.6) / 99.1
(98.5,99.6) / 61.3
(42.2,78.2) / 93.2
(91.8,94.4) / 7.9
(6.6,9.4)
Referral / 417 (21.6) / 54.7
(49.8,59.5)
K / CC follow-up / 323 (85.0) / 69.2
(48.2,85.7) / 89.5
(85.4,92.7) / 36.7
(23.4,51.7) / 97.0
(94.3,98.7) / 8.1
(5.4,11.7)
Referral / 57 (15.0) / 40.4
(27.6,54.2)
L / CC follow-up / 240 (70.8) / 100
(89.7,100) / 14.3
(6.7,25.4) / 38.6
(28.4,49.6) / 100
(66.4,100) / 35.1
(25.6,45.4)
Referral / 99 (29.2) / 52.5
(46.0,59.1)
M / MAU patients
(no CCOS) / 1672 (72.0) / 25.4
(15.8,37.1) / 92.9
(90.5,94.9) / 30.5
(19.2, 43.9) / 91.0
(88.4,93.2) / 10.9
(8.6,13.6)
MAU patients
(with CCOS) / 649 (28.0) / 19.1
(14.0,25.0) / 95.4
(94.2,96.4) / 38.0
(28.8,47.8) / 88.9
(87.2,90.4) / 12.9
(11.3,14.6)
N / CC follow-up / 520 (20.4) / 84.3
(71.4,93.0) / 47.3
(42.7,52.0) / 14.9
(11.0,19.5) / 96.5
(93.2,98.5) / 9.8
(7.4,12.7)
Referral / 2028 (79.6) / 35.5
(33.4,37.6)
O / CC follow-up / 412 (69.6) / 78.0
(65.3,87.7) / 50.4
(41.1,59.7) / 43.8
(34.1,53.8) / 82.2
(71.5,90.2) / 33.1
(26.3,40.6)
Referral / 180 (30.4) / 61.1
(56.2,65.9)
*Includes 1167 (47.4%) CC admissionsnot seen by the CCOS; CC: critical care; CCOS: critical care outreach service; MAU: Medical Admissions Unit; NPV: negative predictive value; PPV: positive predictive value.
Table S3: Results of meta-regression on log diagnostic odds ratio (lnDOR)
Physiological parameters* / Coefficient / 95% confidence intervalTemperature / -1.2 / (-4.4, 1.9)
Urine output / -0.1 / (-1.9, 1.8)
Oxygen saturation / 0.2 / (-1.7, 2.1)
Outcome variables† / Coefficient / 95% confidence interval
Cardiopulmonary resuscitation / 1.2 / (-1.4, 3.9)
Do not attempt resuscitation / 0.7 / (-1.9, 3.4)
Patient group / Coefficient / 95% confidence interval
Critical care follow-up patients‡ / 1.0 / (-1.6, 3.7)
*The following physiological parameters were included in all systems: heart rate, blood pressure, respiratory rate, conscious level; The following physiological parameters were included in a single system only: respiratory support, acid base disturbance;†The following outcomes were recorded in all datasets: admission to critical care, death; ‡Comparison group: all Medical Admissions Unit patients.
1
Figure S1: Receiver operator characteristic (ROC) curves for composite outcome in critical care follow-up patients
1
Figure S2: Forest plot of logdiagnostic odds ratio (lnDOR)
The size of each square is inversely proportional to the variance of lnDOR.The horizontal lines are 95% confidence intervals for lnDOR.
1