Instrument for objective assessment of appropriateness of surgical bed occupancy: validation studyBMJ 2003; 326 doi: http://dx.doi.org/10.1136/bmj.326.7401.1243 (Published 05 June 2003) Cite this as: BMJ 2003;326:1243
- Afshin Alijani, research fellow1,
- George B Hanna, lecturer1,
- Dorin Ziyaie, research fellow1,
- Suzanne L Burns, specialist registrar2,
- Kenneth L Campbell, consultant surgeon1,
- Marion E T McMurdo, professor of ageing and health2,
- Alfred Cuschieri (), head1
- 1 Department of Surgery and Molecular Oncology, Ninewells Hospital and Medical School, Dundee DD1 9SY,
- 2 Department of Medicine, Ninewells Hospital and Medical School
- Correspondence to: A Cuschieri
- Accepted 17 March 2003
The number of hospital beds per head of population has fallen by more than 2% a year since 1980.1 Levels of bed occupancy have risen during this period.1 Appropriate bed occupancy is crucial to the NHS, as occupancy rates exceeding 85% in acute hospitals are associated with problems in dealing with emergency and elective admissions.1 We aimed to develop a valid instrument for the objective assessment of appropriateness of occupancy of surgical beds.
Participants, methods, and results
We drew potential criteria for the instrument from analyses of the medical notes of 200 general surgical inpatients. We chose criteria such that if any one was met we could judge that particular day of surgical care to be appropriate. We shortlisted the criteria after each had been scrutinised by a consultant physician or surgeon of the relevant subspecialty, and we structured them into an instrument. We modified the instrument (box) after a pilot study of 40 bed days.
We conducted a validation study on 100 bed days in a general surgical unit with an interest in colorectal surgery. We selected 10 bed days once a week for 10 weeks by using computer generated random numbers. Two research registrars and a committee of three consultant surgeons in charge of the unit assessed patients independently on the same day. The registrars used the instrument separately to assess the appropriateness of patients' stay; they obtained data from medical notes, charts, history, and clinical examination. The consultants recorded their collective decision, reached by consensus, during a grand ward round. They were blind to instrument criteria and the outcome of registrars' assessments. A geriatrician assessed the elderly patients (> 65 years) who were judged by the instrument to have stayed inappropriately and determined the best placement for the patients on the basis of preset criteria on discharge destination.
We used the degree of agreement between the registrars to measure the interassessor reliability of the instrument. The resistrars reached a consensus on cases about which they disagreed. We measured the predictive clinical validity of the instrument by using the degree of agreement between the registrars' assessment using the instrument and the consultants' collective decision (as the gold standard for valid discharge). We used the assessment by the geriatrician as a control measure.
We studied 100 patients (43 male; 55 emergency, 45 elective), with a median age of 70 (interquartile range 55-81) years. Seventy seven patients had undergone surgery during that admission. The instrument determined that 31% (95% confidence interval 23% to 41%) of patients were occupying beds inappropriately. Of these patients, 26/31–84% (67% to 93%)–were elderly. The instrument had a reliability of 89% (81% to 94%; κ=0.75), a validity of 88% (80% to 93%; κ=0.69), a sensitivity of 93% (84% to 97%), and a specificity of 79% (62% to 89%). We found no pattern of agreement or disagreement between consultants and registrars. The geriatrician agreed with the instrument's assessment in 23/26–88% (71% to 96%)–of cases. The discharge destinations recommended by the geriatrician were to a rehabilitation ward (eight patients), home with extra support (seven), nursing home (three), NHS continuing care (two), medicine for the elderly ward (one), hospice (one), and medical ward (one).
Instrument for objective assessment of appropriateness of surgical bed occupancy
If any one of the following criteria is met, that particular day of care is judged appropriate
Persistent abdominal pain not relieved by oral analgesia
Persistent nausea or vomiting
Abdominal tenderness on palpation, not related to a healthy surgical wound
Intravenous or intramuscular treatment (fluid, feed, drug)
Inhalation treatment (for example, oxygen) not normally being received at home
Inability to move bowels after bowel surgery
Urinary catheterisation for acute retention or urine output measurement
Within 24 hours before day of review
Persistent tachycardia (> 90/min)
Persistent tachypnoea (> 20/min)
Fever > 38°C
Glasgow coma scale ≤ 13/15, acute confusional state
White cell count > 12 or < 4
Evidence of acute renal insufficiency (urine output < 30 ml/hr, creatinine > 130, or urea > 10)
Blood transfusion for acute blood loss
Within 48 hours before day of review
Evidence of acute cardiac ischaemia or arrhythmia
Procedure in operating room that day, not suitable for day surgery unit
Scheduled for an invasive therapeutic or investigative procedure that day requiring inpatient assessment or monitoring
Scheduled for procedure in operating room the next day requiring preoperative consultation or evaluation
24 hours after a procedure, not suitable as day case
Treatment requiring frequent dose adjustments under medical supervision
Monitoring vital signs on surgeon's request, four hourly or more often, for at least 12 hours
Care of major surgical wound, fistula, or drainage
Teaching patient independent stoma care Regardless of the above criteria, a patient is on the ward inappropriately if he or she is
For terminal care or no further active surgical management and
Transferable (physically and ethically)
The study confirmed the reliability and validity of the instrument for the objective evaluation of the appropriateness of surgical bed occupancy in a colorectal unit. A Canadian report on the American appropriateness evaluation protocol showed a poor validity against a panel of experts (κ=0.25).2 A European version of the protocol has been suggested but has not been validated in a substantial study.3 The validity of our instrument needs to be retested in other settings with a larger study population. Whether the instrument can be used as a tool for clinical decision making or audit needs to be confirmed in future studies.
We thank R J C Steele and M Lavelle-Jones for assisting us with conducting this study.
Contributors: GBH had the original idea. AA, DZ, SLB, and KLC conducted the study. AA and GBH wrote the first draft of the paper, which was edited by all the other authors. All authors were involved in the design of the study. AC is the guarantor for the study.
Competing interests None declared.
Ethical approval The local ethics committee approved the study.