Abstract
Background: The Society for Acute Medicine’s Benchmarking Audit (SAMBA) annually examines Clinical Quality Indicators (CQIs) of the care of patients admitted to UK hospitals as medical emergencies.
Aim: The aim of this study is to review the impact of consultant specialty on discharge decisions in the SAMBA data-set.
Design and methods: Prospective audit of patients admitted to acute medical units (AMUs) on 25 June 2015 to participating hospitals throughout the UK with subgroup analysis.
Results: Eighty-three units submitted patient data from 3138 patients.Nearly 1845 (58%, IQR for units 50–69%) of patients were referrals from Emergency Medicine, 1072 (32%, IQR for units 24–44%) were referrals from Primary Care. The mean age was 65 (SD 20). One hundred and forty-one (4.5%) patients were admitted from care homes and 951 (30%) of patients were at least ‘mildly frail’ and 407 (13%) had signs of physiological instability. The median and the mean time to being seen by a doctor were 1 h 20 min and 2 h 3 min, respectively. The median and the mean time to being seen by senior specialist were 3 h 55 min and 5 h 56 min, respectively. By 72 h, 29 (1%) patients had died in the AMU, 73 were admitted to critical care units, 1297 (41%) had been discharged to their own home and 60 to nursing or residential homes. For every 100 patients seen specialists in acute medicine discharged 12 more patients than specialists from other disciplines of medicine (P < 0.001). The difference remained significant after adjustment for case mix.
Conclusion: Specialist in acute care might facilitate discharge in a higher proportion of patients.
Background
Emergency admissions to Acute Secondary Care Hospitals across the United Kingdom (UK) have risen by 47% over the last 15 years.1 The majority of UK hospitals now provide initial diagnoses and treatments, including a review by specialists in internal medicine, in dedicated acute medical units (AMUs). These units receive direct referrals from primary care physicians as well as patients who have been already triaged in the hospital’s emergency departments, which are separate entities under the care of autonomous specialist emergency physicians. AMUs are managed by a mixture of specialists in internal medicine including acute physicians with dedicated post-graduate qualifications in acute medicine.
Despite the well-publicized pressures on European hospitals2 and the standards of care patients should reasonably expect during the admission process and what represents ‘acceptable’ or ‘excellent’ care are poorly defined, and often only confined to a specific disease. Hospital performance can be affected by a number of factors including case-mix and staffing. While the nurse to patient ratio is considered a performance indicator of patient safety,3,4 there is no recognized agreement on the minimum number of doctors required for safe acute medical care.5
Registries for Quality Improvement have the potential to drive performance and support research,6 especially if combined with Patient Related Outcome Measures7 but are often disease specific.8
The Society for Acute Medicine’s Benchmarking Audit (SAMBA) allows units to compare their performance against National Clinical Quality Indicators (CQIs).7–10 To capture case-mix, the audit uses a measure of severity of illness with the National Early Warning Score (NEWS)10 and a measure of frailty (Clinical Frailty Scale (CFS)).11 NEWS correlates with prognosis12 and length of stay in hospital.13 Patients who are frailer have a higher risk of death and significant physiological abnormalities.12
SAMBA’15 was the fourth national audit performed by the Society for Acute Medicine. In addition to National CQIs, this audit also examined specialty input into the AMU patient care and patient experience using the Friends and Family tool.14 Data from the patient feedback have been published separately.15
Methods
Recruitment of centres
All UK hospitals participating in an acute ‘unselected’ medical take were invited to participate in SAMBA’15 and register using a brief online survey.
Patient selection
Inclusion: Patients aged 16 years or above who were seen for admission or assessment on an AMU over a 24-h period from midnight to midnight on 25 June 2015. For hospitals where the AMU is a virtual space in the Emergency Department (ED), with Acute Physicians operating side-by-side with Emergency Physicians, care was only audited on patients who were referred to internal medicine for an inpatient opinion and not on those referred for an urgent outpatient appointment.
Exclusion: Elective patients or day-case patients for technical procedures, such as endoscopies or biopsies.
Data collection
Sites were advised to collect patient data within 12 h of admission, using both clinical records and local patient administration systems. These data were submitted via an online portal, no patient identifiers were included and it contained four sections:
AMU descriptors and staffing levels on the day of the audit. Staffing for the junior and the senior on-call team was measured at 11:00 am, 7:00 pm and 3:00 am.
Patient characteristics: age was reported by decade to assure patient anonymity, vital signs as NEWS values,10 and dependency using the CFS.11
Performance indicators: compliance with National CQIs, access to diagnostics, therapeutics and specialist input during the first 72 h in the AMU.
Patient disposition after 72 h in hospital.
Time zero was defined as the time of admission to hospital (via the emergency department, AMU or other ports of entry). Patient feedback was collected through a separate database in order to prevent identification of individuals.
Summary reports of results were fed back to unit as a comparison with all participating units (sample report, Figure 2).
Figure 2.
Sample scorecard from a participating unit comparing performance with the national data set for all patients submitted (middle row) and for patients admitted directly to the acute medical units (bottom row).
Standards surveyed
The audit tested compliance with the National Acute Medicine CQIs9:
All AMU patients should have their vital signs measured and an early warning score recorded within 30 min of arrival.
All AMU patients should have a full assessment and an appropriate management plan instigated by a competent clinical decision maker within 4 h of arrival. A competent decision maker was defined as a doctor with at least 1 year of experience or a similarly trained health care professional.
All AMU patients should be reviewed by fully trained appropriate specialist (i.e. consultant physician) within 14 h of arrival on the AMU.
Ambulatory care was defined as emergency day care for unscheduled admissions.
Ethics
Ethics approval was waived by the North-West Wales Ethics Committee as data items requested were within national audit guidelines.
Funding
No dedicated funding was received for this audit.
Statistics
Statistics were calculated using SPSS Statistics Version 22.
Results
Unit characteristics
Participating AMUs had a mean of 38 (SD 15) beds with 18 (SD 14) additional beds on associated short-stay wards. Of these, a mean of 1 (SD 3) bed had capability to deliver single organ support (i.e. non-invasive ventilation, inotropic support, etc.). Ambulatory care was delivered by 79 AMUs. In 66 AMUs, ambulatory care was delivered in a dedicated area with a capacity for a mean of 7 (SD 5) patients.
Thirty hospitals had dedicated frailty units for elderly patients, of these, 23 were separate from the AMU.
Staffing of the multi-disciplinary team
On an average, AMUs were staffed from 9 am to 5 pm by a senior doctor and four doctors in training supported by a ward manager, a ward sister, six registered nurses, one pharmacist, one physio-therapist, one occupational therapist, but no social worker. At night, there was an average of five registered nurses working. Health care assistants were also available: 5, 4 and 3 in the morning, evening and night shifts, respectively.
On-call teams, which could be called for assistance at any time, consisted of a median of one registrar (a senior trainee with a post-graduate qualification), two senior house officers (doctors with at least 1 year of experience) and one Foundation Year 1 Trainee (first year after graduation) (Table 1).
Table 1.
Medical staffing of AMUs
| All units | Hospitals with more than 500 beds | Hospitals with 250– 499 beds | |
|---|---|---|---|
| 83 units | 32 | 41 | |
| AMU beds | 37 (15) | 45 (16) | 34 (30) |
| Short-stay units | 17 (14) | 19 (14) | 17 (20) |
| SPR 3:00 h | 1 (0.3) | 1.1 (0.3) | 1.0 (0.2) |
| SPR 11:00 h | 1.4 (0.7) | 1.7 (0.9) | 1.2 (0.5) |
| SPR 19:00 h | 1.3 (0.5) | 1.6 (0.6) | 1.1 (0.4) |
| CMT 3:00 h | 1.8 (0.8) | 2.1 (0.8) | 1.5 (0.7) |
| CMT 11:00 h | 2.5 (1.5) | 2.7 (1.8) | 2.4 (1.5) |
| CMT 19:00 h | 2.5 (1.1) | 3.1 (1.3) | 2.2 (0.9) |
| FY1 3:00 h | 0.6 (0.6) | 0.7 (0.6) | 0.5 (0.5) |
| FY1 11:00 h | 1.3 (0.8) | 1.6 (0.8) | 1.1 (0.8) |
| FY1 19:00 h | 1.4 (0.9) | 1.8 (1.1) | 1.2 (0.8) |
| AMU consultant | 1.6 (1.6) | 1.8 (1.7) | 1.6 (1.6) |
| AMU SPR | 0.9 (0.9) | 1.0 (1.0) | 0.8 (0.9) |
| AMU CMT | 2.0 (1.8) | 2.4 (1.8) | 1.7 (1.7) |
| AMU FY1 | 1.1 (1.4) | 1.1 (1.4) | 1.1 (1.4) |
AMU, acute medical unit; SPR, specialist registrar; CMT, core medical trainee; FY1, foundation year 1 trainee.
Patient cohort 25 June 2015
Eighty-two units submitted data on 3138 patients. Patient characteristics are summarized in Table 1. About 1845 patients (58.8%) were referred from departments of emergency medicine and 1072 (34.2%) from general practitioners in primary care (Figure 1).
Figure 1.
Sources of admission to Acute Medical Units.
The median number of AMU admissions during the study period was 36 patients (IQR 26–47). The mean age was 65 years (SD 20, median 70). Characteristics are summarized in Table 2. Of the 2991 patients for whom deposition was known, 1357 (43%) patients were discharged directly from the AMU and 1458 (46%) patients were transferred to another area of care. The mean time to discharge home was 14 h and the mean time to transfer to a specialty ward was 17 h.
Table 2.
Patient characteristics
| Characteristics | N (%) | Comments |
|---|---|---|
|
Gender |
Female 1599 (51%) |
|
|
|
|
| Dementia | 235 patients (7.5%) | |
| Delirium | 263 patients (8.4%) | |
| Physiological instability (NEWS > =5) | 407 (13%) | IQR for units 8-17% |
|
|
|
|
|
|
| Readmission in previous 28 days for same or another condition | 13% (IQR 9–17%) | |
|
|
Usage of ambulatory care
About 428 (14%) patients were seen first in the ambulatory care. Of the patients seen in ambulatory care, only 13 had a NEWS of more than four and 59 a CFS of more than four and only four patients admitted to ambulatory care had both a NEWS above 4 and a CFS above 4.
Clinical quality indicators
About 73% of patients had a NEWS within 30 min of hospital arrival (i.e. either to ED or AMU). For patients who were directly admitted to the AMU compliance with this CQI was marginally lower at 71%. The median and the mean time to first NEWS were 10 min and 26 min, respectively.
About 87% of all patients were seen by a doctor within 4 h of arrival, and 91% of patients who were admitted directly to the AMU. The median and mean times to being seen were 1 h 20 min and 2 h 3 min, respectively.
Although 78% of all patients’ were seen by a consultant within 14 h of arrival, only 71% of those admitted directly to the AMU were seen within this time frame. The median and the mean time to being seen were 3 h 55 min and 5 h 56 min, respectively.
Approximately 52% of patients met all three of the National CQIs and 47% of those admitted directly to the AMU. Fourteen (17%) units achieved all three indicators in over 70% of patients admitted. Summaries of unit data showed broad variation for all indicators (Figure 2).
Compliance with CQIs was worst in the evening shift – 33% of patients were compliant with all three indicators compared with 47% seen between 9 am and 5 pm and 60% during the night shift. This was often as a result of the delayed senior review of patients admitted in the evening: timely senior review was found in 74.1% of patients admitted during day-time, 55.2% during evening hours and in 92.7% of patients during night-time.
Specialist input
The first senior review after admission (‘Post-Take Ward Round’) patients was conducted by a consultant specialist in acute medicine in 1911 (61%) patients, by a geriatrician in 252 (8.0%), a chest physician in 158 (5.0%), an endocrinologist in 141 (4.4%), a gastroenterologist in 114 (3.6%), a cardiologist, in 76 (2.4%) and by a nephrologist in 73 (2.2) patients. Other specialties accounted for 7%. In 13 units, 90% or more of the patients were seen by acute physicians, and in 6 units, 100% of patients were seen by acute physicians.
Patients who were seen by a specialist in acute medicine were more likely to return to their own home within 72 h (845/1886 (45%) vs. 342/1033 (33%), P < 0.0001). Discharge rates were higher in patients with a lower level of physiological derangement as measured by a NEWS of less than 5 (761/1585 (48%) vs. 308/864 (36%), P < 0.0001) and in patients with a low degree of frailty as measured by a Clinical Frailty Scale of less than 5 (727/1320 (55%) vs. 275/652 (42%), P < 0.0001). Binary logistic regression using referral by the emergency department, admission through the emergency department, presence of dementia, presence of acute confusion, gender, NEWS of less than 5, Frailty of less than 5 and readmission as variable suggested that a review by a specialist in acute medicine was associated with a higher discharge rate. For every 100 patients seen specialists in acute medicine discharged 12 more patients than specialists from other disciplines of medicine (P < 0.001).
Following initial senior review, 884 patients were referred to a number of sub-specialties (Table 3). Cardiology specialist input accounted for the largest number with 198 referrals. Sub-specialty input was usually provided on the next working day.
Table 3.
Specialty input in acute medicine: time from referral to review
| Medical specialties | n | Hours from admission to review (median) |
|---|---|---|
| Cardiology | 198 | 14:28 |
| Gastro-enterology | 94 | 16:00 |
| Respiratory | 86 | 16:44 |
| Geriatrics | 83 | 16:29 |
| Neurology | 25 | 16:18 |
| Endocrinology | 23 | 14:20 |
| Nephrology | 16 | 15:02 |
| Oncology | 1 | n.a. |
| Other specialties | ||
| Orthopaedics | 22 | 7:02 |
| General Surgery | 16 | 6:31 |
| Neuro-surgery | 3 | n.a. |
| Psychiatry | 35 | 12:37 |
| Other | 282 | n.a. |
Discussion
SAMBA’15 shows performance of the front door of UK departments for internal medicine: severity of illness scores were being measured early on admission, four out of five patients saw a doctor within the recommended time, and an ever-increasing proportion or patients were being seen by dedicated specialists in acute medicine. Importantly specialists in acute medicine achieved higher rates of early discharge even in a population where a third of all patients were measurably frail: for every 100 patients seen this resulted in 12 patients being discharged early. To our knowledge, the impact of speciality on performance in general internal medicine has not been previously investigated in a multi-centre study.
SAMBA’15 is only a one-day snapshot and might not, therefore, represent the best or typical performance of the participating units. For example, one of the hospitals had only opened 9 days before the audit took place. While in a small proportion of units, all or nearly all patients are seen by acute physicians, the number of unfilled consultant posts and the availability of training posts limits current impact. Disappointingly performance against CQIs has not greatly improved since 2013. A possible cause is the mismatch between staff working in the early evening hours and the number of patients being admitted during this time period. Re-calibration of shifts with a higher proportion of doctors working in evenings can lead to less waiting for patients, better performance and arguably a better patient experience.16
A quarter of all SAMBA patients were referred in the AMU to other specialties, more than 500 to medical sub-specialties. In a time where pressure on bed capacity is rising, it would seem sensible to focus on early senior review with responsive input from physicians from sub-specialties where needed. This might require a cooperative model of specialist care outside traditional silos.
The number of patients that are seen in dedicated ambulatory care units remains small (14%) and below the expectations of the ambulatory care networks which claims that about a third of patients might be able to quality for this type of ‘bed-less’ care17 to provide timely cost-effective and patient centred care. At the same time, ambulatory care might be able to capture what in some international studies have been labelled as ‘inappropriate hospital admissions’18 and allow a more comprehensive comparison of cost drivers between systems. We are aware that some units run speciality-specific ambulatory care services that might not have been captured as part of SAMBA.
The current SAMBA data set collects data on staffing levels, but the significant variations in shift patterns, inter-disciplinary working and the way in which hospital admission processes are configured have made it difficult in the past to determine which models of service delivery are likely to achieve the best quality of care.12 The enormous variation in hospital configuration makes comparisons difficult. A system to classify typical configurations of hospitals would help those working in similar units to learn from each other. Although how to measure quality may be debatable, it must the right to at least agree to start the clock for all patients at the time of hospital admission, and not at some arbitrary time thereafter.19
The relationship between emergency medicine and acute internal medicine requires on-going and further evaluation: it is obvious that many patients experience double handling and that the degree of synchronization of services between EDs and AMUs will affect both patient experience and other quality indicators.
Conclusion
SAMBA’15 gives a clear indication about the gap between the ambition to deliver persistent and the sustainable high-quality care in UK hospitals and the difficulties to align resources in some units to achieve this. The SAMBA model of measurement is applicable nationally and potentially internationally and could be used to underpin evaluation of innovative service models that can benefit patients throughout Europe.
Supplementary Material
Acknowledgements
The authors would like to thank the Research Committee of the Society for Acute Medicine in General and the Chair, Dr Louella Vaughan in particular for facilitating, hosting and promoting the Benchmarking Audit and for contributions to the format of both the audit and the resulting publications. We would like to express our gratitude to Professor Kenneth Rockwood for the permission in using the Clinical Frailty Scale. Sara Pradhan received a Wellcome Trust Vacation Scholarship.
Data collectors
Data for SAMBA ‘15 was collected by Kevin Carter, Jan Droste, Siddhesh Prabhavalkar, Vijdan Gani, Abdul Majeed, Nasir Ameer, Philip Swales, Matthew Maw, Wendy Munro, Tom Payne, Nick Scriven, Hannah Skene, Stephan Birkner, Anthony Darby, Paarul Prinja, Edward Pineles, James Shawcross, Aileen Coupe, Tendekayi Msimang, Carmen Ersek, Michael Dias, Ruth Petch, Muhammad Raza, Susan Fair, Rob Hughes, Ariyur V Balaji, Deblina Dasgupta, Mohandas Dasa, Matthew Giles, Charlotte Masterton-Smith, Kamal Naserl Aram Salehi, Shirley Hammersley, Alex Taylor, Mark Cranston, Rachel Anderson, Alex Keough, Tania A Syed, A Thompson, Katharine Willmer, Adnan Jan, Aylwin Chick, Aled Huws, Joanne Morris, Subash Sivaraman, Ivan Le Jeune, Innes Young, Szabolcsi and Gilchrist, Alasdair Miller, Leslie R Ala, Umed M Nadir, Faisal Mohammad, Zia Din, Andrea Adjetey, Belen Espina, Eleanor Campbell, Hassan Paraiso, Adnan Gebril, Paarul Prinja, Rajesh Thimmappa, Vijaykumar Singh, Shiva Sreenivasan, Chris Roseveare, Joanna Peasegood, Henry Gibson, Jane Evans, Katy Islip, Neal Gent, Tapas Chakraborty, Alice Miller, Gaurav Agarwal, Khin Nyo, Mike Jones, Viviana Elliott, Ethel Gomez Canales, Ragit Varia, Clarissa Murdoch, Bill Lusty, Roger Duckitt, Sarah Dyer, Mark Holland, Tendekayi Msimanga, Belen Espina, Alex Taylor, Neal Gent, Mohandas Dasan, Zia Din, Shirley Hammersley and Anthony Darby.
Deysha Ratnasingham, Stuart Stevenson and many others who’s might have facilitated the data collection but who’s names were not passed on.
Conflict of interest. None declared.
References
- 1. National Audit Office. Emergency admissions to hospital: managing the demand, 2013.
- 2. European Hospital and Healthcare Federation. Hospitals in Europe healthcare data 2011, 2011.
- 3. Aiken LH, Clarke SP, Sloane DM, Sochalski JSJH. Hospital nurse staffing and patient mortality, nurse burnout, and job dissatisfaction. JAMA [Internet] 2002; 288:1987–93. Available from: http://www.nursing.upenn.edu/media/California legislation/Documents/Linda [DOI] [PubMed] [Google Scholar]
- 4. Fenton KCA. A tool to calculate safe nurse staffing levels. Nurs Times 2015; 111:12–4. [PubMed] [Google Scholar]
- 5. Sabin AJ, Subbe Hristian P, Vaughan L, Dowdle R. Safety in numbers: lack of evidence to indicate the number of physicians needed to provide safe acute medical care. Clin Med (Northfield, IL) 2014; 14:462–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Levay C. Policies to foster quality improvement registries: lessons from the Swedish case. J Intern Med 2016; 279:160–72. [DOI] [PubMed] [Google Scholar]
- 7. Nilsson E, Orwelius L, Kristenson M. Patient-reported outcomes in the Swedish National Quality Registers. J Intern Med [Internet] 2015; n/a–n/a. Available from: http://doi.wiley.com/10.1111/joim.12409 [DOI] [PubMed] [Google Scholar]
- 8. Emilsson L, Lindahl B, Köster M, Lambe M, Ludvigsson JF. Review of 103 Swedish Healthcare Quality Registries. J Intern Med (Blackwell Publishing Ltd); 2015; 277:94–136. [DOI] [PubMed] [Google Scholar]
- 9.The Society for Acute Medicine. Clinical Quality Indicators for Acute Medical Units (AMUs). 2011, 2. Available from: http://www.acutemedicine.org.uk/wp-content/uploads/2010/11/clinical_quality_indicators_for_acute_medical_units_v18.pdf
- 10. Jones M. NEWSDIG: the national early warning score development and implementation group. Clin Med, J R Coll Phys Lond 2012; 501–3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Rockwood K, Song X, MacKnight C, Bergman H, Hogan DB, McDowell I, et al. A global clinical measure of fitness and frailty in elderly people. CMAJ 2005; 173:489–95. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Subbe CP, Burford C, Jeune IL, Masterton-Smith C, Ward D. Relationship between input and output in acute medicine – secondary analysis of the Society for Acute Medicine’s benchmarking audit 2013 (SAMBA '13). Clin Med (Northfield, IL) [Internet] 2015; 15:15–9. Available from: http://www.clinmed.rcpjournal.org/content/15/1/15.full\nhttp://www.clinmed.rcpjournal.org/cgi/doi/10.7861/clinmedicine.15-1-15 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Paterson R, MacLeod DC, Thetford D, Beattie A, Graham C, Lam S, et al. Prediction of in-hospital mortality and length of stay using an early warning scoring system: clinical audit. Clin Med J R Coll Phys Lond 2006; 6:281–4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Publications Gateway Ref. No. 01787. The friends and family test: Guidance, 2015.
- 15. Pradhan S, Ratnasingham D, Subbe CP, Stevenson S, Ward D, Cooksley T, et al. Society for acute medicine benchmarking audit 2015: the patient perspective. Acute Med 2015; 14:147–50. [PubMed] [Google Scholar]
- 16. Le Jeune IR, Simmonds MJ, Poole L. Patient need at the heart of workforce planning: the use of supply and demand analysis in a large teaching hospital’s acute medical unit. Clin Med 2012; 12:312–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Institute for Innovation and Improvement. Directory of Ambulatory Emergency Care for Adults, 2010.
- 18. Eriksen BO, Kristiansen IS, Nord E, Pape JF, Almdahl SM, Hensrud A, et al. The cost of inappropriate ad mi ssions: a study of health benefits and resource utilization in a department of internal medicine. J Intern Med 1999; 246:379–87. [DOI] [PubMed] [Google Scholar]
- 19. Lang S, Cooksley T, Foden P, Holland M. Acute medicine targets: when should the clock start and 7-day consultant impact? QJM 2015; 108:611–6. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.