'Weekend Effect': What Is the Evidence?

The 'weekend effect': What is the evidence?

1st June 2016

Paul Aylin Professor of Epidemiology and Public Health [email protected] Imperial Patient Safety Translational Research Centre The ‘weekend effect’

• Background • My work • Explanations • Next steps

Freemantle et al.

• Update of their 2012 paper on 2009/10 admissions for 2013/14 • Compared with Wednesday

• Combines elective and emergency admissions • Choice of disease specific analyses (Oncology and cardiovascular) • Day of death analysis 1. Abougergi MS, Travis AC, Saltzman JR. Impact of day of admission on Increased mortality associated with weekend hospital admission: a case for 68. O'Neill DE, Southern DA, O'Neill BJ, McMurtry MS, Graham MM. Weekend mortality and other outcomes in upper GI hemorrhage: a nationwide analysis. expanded 7 day services? BMJ 2015;351 :h4596 compared with weekday presentation does not affect outcomes of patients Gastrointest Endosc. 2014 Aug; 80(2): 228-35 presenting with non-ST elevation acute coronary syndrome. Eur Heart J 35. Freemantle, Keogh, Pagano et al. Weekend hospitalization and additional http://www.ncbi.nlm.nih.gov/pubmed/24674354 Acute Cardiovasc Care. 2014 Jun; 3(2): 99-104. risk of death: An analysis of inpatient data. N J R Soc Med 2012; 105(2): 74- http://www.ncbi.nlm.nih.gov/pubmed/24585942 2. Albright KC, Savitz SI, Raman R, Martin-Schild S, Broderick J, Ernstrom K, 84 http://www.ncbi.nlm.nih.gov/pubmed/22307037 Ford A, Khatri R, Kleindorfer D, Liebeskind D, Marshall R, Merino JG, Meyer 69. Orandi BJ, Selvarajah S, Orion KC, Lum YW, Perler BA, Abularrage CJ. 36. Gallerani M, Imberti D, Ageno W, Dentali F, Manfredini R. Higher mortality DM, Rost N, Meyer BC. Comprehensive stroke centers and the 'weekend Outcomes of nonelective weekend admissions for lower extremity ischemia. J rate in patients hospitalised for acute pulmonary embolism during weekends. effect': the SPOTRIAS experience. Cerebrovasc Dis. 2012; 34(5-6): 424-9 Vasc Surg. 2014 Dec; 60(6): 1572-9 Thromb Haemost. 2011 Jul; 106(1): 83-9 http://www.ncbi.nlm.nih.gov/pubmed/23207423 http://www.ncbi.nlm.nih.gov/pubmed/25441678 http://www.ncbi.nlm.nih.gov/pubmed/21544321 3. Al-Lawati JA, Al-Zakwani I, Sulaiman K, Al-Habib K, Al Suwaidi J, 70. Orman ES, Hayashi PH, Dellon ES, Gerber DA, Barritt AS 4th. Impact of 37. Gellerani M et al. Aortic rupture. Journal of Vascular Surgery. Volume 55, Panduranga P, Alsheikh-Ali AA, Almahmeed W, Al Faleh H, Al Saif S, Hersi nighttime and weekend liver transplants on graft and patient outcomes. Liver Issue 5 , Pages 1247-1254, May 2012. Higher mortality in patients A, Asaad N, Al-Motarreb A,Over Mikhailidis DP, Amin 100 H. Weekend papers versus Transpl. 2012 May; 18(5): 558-65 hospitalized for acute aortic rupture or dissection during weekends weekday, morning versus evening admission in relationship to mortality in http://www.ncbi.nlm.nih.gov/pubmed/22271668 http://www.ncbi.nlm.nih.gov/pubmed/22542339 acute coronary syndrome patients in 6 middle eastern countries: results from 71. Ozdemir BA, Sinha S, Karthesalingham, Poloniecki J, Pearse RM, Grocott gulf race 2 registry. Open Cardiovasc Med J. 2012; 6: 106-12 38. Goldacre MJ, Maisonneuve JJ. Mortality from meningococcal disease by day MPW, Thompson MM, Holt P. Mortality of emergency general surgical http://www.ncbi.nlm.nih.gov/pubmed/23002404 of the week: English national linked database study J Public Health (Oxf). patients and associations with hospital structures and processes. British 2013 Sep; 35(3): 413-21 http://www.ncbi.nlm.nih.gov/pubmed/23378233 4. Ananthakrishnan AN et al. Outcomes of Weekend Admissions for Upper Journal of Anaesthesia 2016;116 (1):54-62 Gastrointestinal Hemorrhage: A Nationwide Analysis. Clin Gastro Hepatology 39. Goldstein SD, Papandria DJ, Aboagye J, Salazar JH, Van Arendonk K, Al- 72. Palmer W, Bottle A, Aylin P. Association between day of delivery and obstetric 2009; 7(3): 296-302.e1 http://www.ncbi.nlm.nih.gov/pubmed/19084483 Omar K, Ortega G, Sacco Casamassima MG, Abdullah F. The "weekend outcomes: observational study. BMJ 2015;351:h5774 effect" in pediatric surgery - increased mortality for children undergoing urgent 5. Arabi Y, Alshimemeri A, Taher S.. Weekend and weeknight admissions have surgery during the weekend. J Pediatr Surg. 2014 Jul; 49(7): 1087-91. 73. Palmer WL et al. A Retrospective Cohort Study on the Association Between the same outcome of weekday admissions to an intensive care unit with http://www.ncbi.nlm.nih.gov/pubmed/24952794 Day of Hospital Presentation and the Quality and Safety of Stroke Care. Arch onsite intensivist coverage. Crit Care Med. 2006 Mar; 34(3): 605-11 Neurol 2012; 69(10): 1296-1302 http://www.ncbi.nlm.nih.gov/pubmed/16521254 40. Goodman EK, Reilly AF, Fisher BT, Fitzgerald J, Li Y, Seif AE, Huang YS, http://www.ncbi.nlm.nih.gov/pubmed/22777008 Bagatell R, Aplenc R. Association of weekend admission with hospital length 6. Aylin P et al. Day of week of procedure and 30 day mortality for elective of stay, time to chemotherapy, and risk for respiratory failure in pediatric 74. Pasupathy D, Wood AM, Pell JP, Fleming M, Smith GC. Time of birth and risk surgery: retrospective analysis of hospital episode statistics.. BMJ 2013; 346: patients with newly diagnosed leukemia at freestanding US children's of neonatal death at term: retrospective cohort study. BMJ 2010;341:c3498. f2424 http://www.ncbi.nlm.nih.gov/pubmed/23716356 hospitals. JAMA Pediatr. 2014 Oct; 168(10): 925-31 75. Pederby MA et al. Survival From In-Hospital Cardiac Arrest During Nights 7. Aylin P et al. Weekend mortality for emergency admissions. A large, http://www.ncbi.nlm.nih.gov/pubmed/25155012 and Weekends. JAMA 2008; 299(7): 785-792 multicentre study. Qual Saf Health Care 2010; 19(3): 213-217 41. Groves EM, Khoshchehreh M, Le C, Malik S. Effects of weekend admission http://www.ncbi.nlm.nih.gov/pubmed/18285590 http://www.ncbi.nlm.nih.gov/pubmed/20110288 on the outcomes and management of ruptured aortic aneurysms. J Vasc 76. Powell ES, Khare RK, Courtney DM, Feinglass J. The weekend effect for 8. Barba R et al. The impact of weekends on outcome for acute exacerbations of Surg. 2014 Aug; 60(2): 318-24 patients with sepsis presenting to the emergency department . J Emerg Med. COPD. ERJ 2012; 39(1): 46-50 http://www.ncbi.nlm.nih.gov/pubmed/24709439 2013 Nov;45(5):641-8. doi: 10.1016/j.jemermed.2013.04.042. Epub 2013 Aug http://www.ncbi.nlm.nih.gov/pubmed/21659418 42. Hamaguchi S, Kinugawa S, Tsuchihashi-Makaya M, Goto D, Tsutsui H. 30. http://www.ncbi.nlm.nih.gov/pubmed/23993937 9. Barnett MJ et al. Day of the week of intensive care admission and patient Weekend versus weekday hospital admission and outcomes during 77. Ricciardi R, Nelson J, Roberts PL, Marcello PW, Read TE, Schoetz DJ. Is the outcomes: a multisite regional evaluation. Med Care 2002; 40(6): 530-9 hospitalization for patients due to worsening heart failure: a report from presence of medical trainees associated with increased mortality with http://www.ncbi.nlm.nih.gov/pubmed/12021679 Japanese Cardiac Registry of Heart Failure in Cardiology (JCARE-CARD). weekend admission? BMC Med Educ. 2014 Jan 8;14:4. doi: 10.1186/1472- Heart Vessels. 2014 May; 29(3): 328-35 10. Bejanyan N, Fu AZ, Lazaryan A, Fu R, Kalaycio M, Advani A, Sobecks R, 6920-14-4 http://www.ncbi.nlm.nih.gov/pubmed/24397268 http://www.ncbi.nlm.nih.gov/pubmed/23653107 Copelan E, Maciejewski JP, Sekeres MA. Impact of weekend admissions on 78. Robinson E, Smith G, Power S, Harrison D, Nolan J, Soar J, Spearpoint K, quality of care and outcomes in patients with acute myeloid leukemia Cancer. 43. Hamilton P et al. Weekend Birth and Higher Neonatal Mortality: A Problem of Gwinnutt C, Rowan K. Risk-adjusted survival for adults following in-hospital 2010 Aug 1; 116(15): 3614-20 http://www.ncbi.nlm.nih.gov/pubmed/20564070 Patient Acuity or Quality of Care? Journal of Obstetric, Gynecologic, & cardiac arrest by day of week and time of day: observational cohort study. Neonatal Nursing 2003; 32(6): 724–733 11. Béjot Y, Aboa-Eboulé C, Jacquin A, Troisgros O, Hervieu M, Durier J, Osseby BMJ Qual Saf 2015. http://dx.d http://www.ncbi.nlm.nih.gov/pubmed/14649592 GV, Giroud M. Eur J Neurol. Stroke care organization overcomes the 79. Ruiz M, Bottle A, Aylin P. The Global Comparators Project: international deleterious 'weekend effect' on 1-month stroke mortality: a population-based 44. Handel EH et al. Weekend admissions as an independent predictor of comparison of 30 day in-hospital mortality by day of the week. BMJ Quality study. 2013 Aug; 20(8): 1177-83 mortality: an analysis of Scottish hospital admissions. BMJ Open 2012; 2: and Safety 2015;0:1-13 doi:10.1136/bmjqs-2014-003467 http://www.ncbi.nlm.nih.gov/pubmed/23551852 e001789 http://www.ncbi.nlm.nih.gov/pubmed/23135542 80. Sakhuja A et al. Outcomes of Patients Receiving Maintenance Dialysis 12. Bell CM, Redelmeier DA. Mortality among patients admitted to hospitals on 45. Hansen KW, Hvelplund A, Abildstrøm SZ, Prescott E, Madsen M, Madsen JK, Admitted Over Weekends. Ankit Sakhuja. AJKD 2013; 62(4): 763-770 weekends as compared with weekdays. N Engl J Med. 2001 Aug 30; 345(9): Jensen JS, Sørensen R, Galatius S. Prognosis and treatment in patients http://www.ncbi.nlm.nih.gov/pubmed/23669002 663-8 http://www.ncbi.nlm.nih.gov/pubmed/11547721 admitted with acute myocardial infarction on weekends and weekdays from 1997 to 2009. Int J Cardiol. 2013 Sep 30; 168(2): 1167-73 81. Salihu HM, Ibrahimou B, August EM, Dagne G. Risk of infant mortality with 13. Bell D, Lambourne A, Percival F, Laverty AA, Ward DK. Consultant input in http://www.ncbi.nlm.nih.gov/pubmed/23199552 weekend versus weekday births: a population-based study. J Obstet acute medical admissions and patient outcomes in hospitals in England: a Gynaecol Res. 2012 Jul; 38(7): 973-9 multivariate analysis PLoS One. 2013 Apr 17; 8(4): e61476 46. Hixson ED, Davis S, Morris S, Harrison AM. Do weekends or evenings matter http://www.ncbi.nlm.nih.gov/pubmed/22487462 http://www.ncbi.nlm.nih.gov/pubmed/23613858 in a pediatric intensive care unit? Pediatr Crit Care Med. 2005 Sep; 6(5): 523- 30. http://www.ncbi.nlm.nih.gov/pubmed/16148810 82. Saposnik G, Baibergenova A, Bayer N, Hachinski V. Weekends: a dangerous 14. Bray B, Cloud GC, James MA, Hemmingway H, Paley L, Stewart K, Tyrrell time for having a stroke? Stroke. 2007 Apr; 38(4): 1211-5 PJ, Wolfe CDA, Rudd AG. Weekly variation in health-care quality by day and 47. Hoh BL, Chi YY, Waters MF, Mocco J, Barker FG. Effect of weekend http://www.ncbi.nlm.nih.gov/pubmed/17347472 time of admission: a nationwide, registry-based, prospective cohort study of compared with weekday stroke admission on thrombolytic use, in-hospital acute stroke care. Lancet 2016; D mortality, discharge disposition, hospital charges, and length of stay in the 83. Schmid M, Ghani KR, Choueiri TK, Sood A, Kapoor V, Abdollah F, Chun FK, Nationwide Inpatient Sample Database, 2002 to 2007. 2nd. Stroke. 2010 Oct; Leow JJ, Olugbade K Jr, Sammon JD, Menon M, Kibel AS, Fisch M, Nguyen 15. Bray BD, Ayis S, Campbell J, Cloud GC, James M, Hoffman A, Tyrrell PJ, 41(10): 2323-8 http://www.ncbi.nlm.nih.gov/pubmed/20724715 PL, Trinh QD. An evaluation of the 'weekend effect' in patients admitted with Wolfe CD, Rudd AG. Associations between stroke mortality and weekend metastatic prostate cancer. BJU Int. 2014 Aug 7. doi: 10.1111/bju.12891. working by stroke specialist physicians and registered nurses: prospective 48. Hong JS et al. Comparison of Case Fatality Rates for Acute Myocardial [Epub ahead of print] http://www.ncbi.nlm.nih.gov/pubmed/25099032 multicentre cohort study. PLoS Med. 2014 Aug 19; 11(8) Infarction in Weekday vs Weekend Admissions in South Korea. Circulation http://www.ncbi.nlm.nih.gov/pubmed/25137386?dopt=Abstract Journal 2010; 74(3): 496-502 http://www.ncbi.nlm.nih.gov/pubmed/20075558 84. Schmulewitz L, Proudfoot A, Bell D. The impact of weekends on outcome for emergency patients. Clin Med 2005;5:621e5 16. Brims FJ, Asiimwe A, Andrews NP, Prytherch D, Higgins BR, Kilburn S, 49. Horwich TB et al. Weekend hospital admission and discharge for heart failure: Chauhan AJ. Weekend admission and mortality from acute exacerbations of Association with quality of care and clinical outcomes. American Heart 85. Schneider EB et al. Beating the weekend trend: Increased mortality in older chronic obstructive pulmonary disease in winter. Clin Med. 2011 Aug; 11(4): Journal 2009; 158(3): 451-458 adult traumatic brain injury (TBI) patients admitted on weekends. J Surg Res 334-9 http://www.ncbi.nlm.nih.gov/pubmed/21853828 http://www.ncbi.nlm.nih.gov/pubmed/19699870 2012; 177(2): 295-300 http://www.ncbi.nlm.nih.gov/pubmed/22795343 17. Brims FJ, Asiimwe A, Andrews NP, Prytherch D, Higgins BR, Kilburn S, 50. Ibrahimou B, Salihu HM, English G, Anozie C, Lartey G, Dagne G. Twins born 86. Shaheen AA, Kaplan GG, Myers RP. Clin Gastroenterol Hepatol . Weekend Chauhan AJ. Weekend admission and mortality from acute exacerbations of over weekends: are they at risk for elevated infant mortality? Arch Gynecol versus weekday admission and mortality from gastrointestinal hemorrhage chronic obstructive pulmonary disease in winter. Clin Med. 2011 Aug; 11(4): Obstet. 2012 Dec; 286(6): 1349-55 caused by peptic ulcer disease. 2009 Mar; 7(3): 303-10 Bell et al. NEJM 2001

• Mortality among patients admitted to hospital on weekends as compared with weekdays Canadian study, acute care from ER departments in Ontario, Canada. • 3.8 million admissions • Looked at specific conditions (AAA, acute epiglottis, PE, AMI, Stroke and #NOF) plus the 100 conditions that were the most common causes of death. • Significant weekend effect for AAA (OR 1.28), epiglottis (OR 5.47) and PE (OR 1.19), plus 23 out of the 100 leading causes of death. • No conditions associated with significantly lower mortality rates at the weekend.

Bell CM, Redelmeier DA. Mortality among patients admitted to hospital on weekends as compared with weekdays. N Engl J Med 2001;345:663e8. Kostis et al. 2007

• Weekend versus weekday admission and mortality from myocardial infarction • Acute Myocardial Infarction, all admissions in New Jersey 1987- 2002 • 231,164 admissions • In the interval from 1999 to 2002 (59,786 admissions), mortality at 30 days was significantly higher for patients admitted on weekends (12.9% vs. 12.0%, P = 0.006). • Persisted at 1 year (1% absolute difference in mortality). • Associated with lower rate of invasive procedures

Kostis WJ, Demissie K, Marcella SW, et al. Weekend versus weekday admission and mortality from myocardial infarction. N Engl J Med 2007;356:1099e109 Barba et al. 2006

• Mortality among adult patients admitted to the hospital on weekends • All acute admissions to tertiary acute care hospital in Spain (1999-2003) • 35,993 hospital admissions • All in-hospital deaths OR = 1.1, but not significant. Deaths within 48 hours OR=1.4

Barba R, Losa JE, Velasco M, et al. Mortality among adult patients admitted to the hospital on weekends. Eur J Intern Med 2006;17:322e4

Imperial Patient Safety Translational Research Centre Other studies

• Hamilton et al found higher weekend neonatal mortality (OR= 1.42) in Texas 1999-2000. • Schmulewitz et al found no weekend effect in a single Scottish hospital • Wunsch et al used ICNARC to examine 75,621 admissions to 102 ICUs across England, Wales and NI and found admissions on weekend days (Friday, Saturday, Sunday) or in the evening/night were associated with higher odds of crude hospital death, but this was accounted for by case mix.

Hamilton P, Resttrepo E. Weekend birth and higher neonatal mortality: a problem of patient acuity or quality? J Obstet Gynecol Neonatal Nurs 2003;32:724e33. Schmulewitz L, Proudfoot A, Bell D. The impact of weekends on outcome for emergency patients. Clin Med 2005;5:621e5. Wunsch H, Mapstone J, Brady T, et al. Hospital mortality associated with day and time of admission to intensive care units. Intensive Care Med 2004;30:895e901. Literature

• Studies finding no effect tend to be smaller, single hospital studies or based in intensive care Mikulich et al

• St. James Hospital, Dublin • 49,337 episodes of inpatient care • 30 day in-hospital mortality at the weekend (9.99% vs. 9.0%) • Wide confidence intervals (small sample) • Unadjusted OR 1.11 (CI 0.99-1.23) • Adjusted OR 1.05 (CI 0.88-1.24) • Entirely consistent with other papers showing weekend effects

Weekend mortality for emergency admissions

• All emergency admissions in England for 2005/6 • Examined top 50 diagnoses leading to death and all cause admissions. • Adjusted for age, sex, socioeconomic deprivation, comorbidity and diagnosis • Odds of death were calculated for admissions at the weekend compared to admissions during the week

Imperial Patient Safety Translational Research Centre Weekend mortality for emergency admissions

• 4,317,866 admissions with 215,054 in-hospital deaths • Crude mortality • 5.2% for weekend admissions • 4.9% for weekday admissions • 17 out of 50 diagnosis groups associated with significantly higher odds of death (p<0.001) including stroke (OR 1.13), AMI (OR1.08) • Overall adjusted odds of death was 10% higher (OR=1.10, 95% CI 1.08-1.11) in those patients admitted at the weekend compared with patients admitted during a weekday (p<0.001) Results Limitations

• Multiple statistical tests • No out of hospital deaths • Fewer admissions at weekend, so casemix might be different

Methods

• All acute and specialist English hospitals carrying out elective surgery over three financial years, from 2008-09 to 2010-11 • Planned operating room planned procedures based on AHRQ patient safety indicator1 • Exclude day cases • 30 day post op deaths (including out of hospital deaths linked to ONS death data) • Adjusted for age, sex, ethnic group, socio-economic deprivation, co-morbidities, number of emergency admissions in the last 12 months, year and procedure risk quintile

1 Bottle A, Aylin P, 2009, Application of AHRQ patient safety indicators to English hospital data, Quality & Safety in Health Care, Vol:18, ISSN:1475-3898, Pages:303-308 Results

• 27,582 deaths within 30 days following 4,133,346 inpatient admissions for elective operating room procedures (6.7 per 1000). • Compared with Monday, adjusted odds of death were 44% and 82% higher, respectively, if the procedures were carried out on Friday (odds ratio 1.44, 95% confidence interval 1.39 to 1.50) or a weekend (1.82, 1.71 to 1.94)

1 Bottle A, Aylin P, 2009, Application of AHRQ patient safety indicators to English hospital data, Quality & Safety in Health Care, Vol:18, ISSN:1475-3898, Pages:303-308 Adjusted odds of death and 95% confidence intervals by day of procedure in English hospitals for 2008-9 to 2010-11. (Crude rate 0.67%)

(exc. day cases and Aylin P, Alexandrescu R, Jen MH, et al., 2013, Day of week of procedure and 30 day mortality for elective surgery:deliveries) retrospective analysis of hospital episode statistics, British Medical Journal, Vol:346, ISSN:1756-1833, Pages:f2424-f2424 ©2013 by British Medical Journal Publishing Group Discussion

• Not able to use more recent data • Mortality low for elective procedures (6.7 per 1000) • Monday benchmark • Hypothesis that post-op care may have contributed to mortality • Same picture in 2 day mortality • Fall in Sunday mortality Adjusted odds of death and 95% confidence intervals by day of procedure in English hospitals for 2008-9 to 2010-11. 2.2

2.0

1.8 OR and 95% CI 95% and OR 1.6

1.4

1.2

1.0

0.8 Mon Tues Wed Thu Fri Sat Sun Days of week

(exc. day cases and deliveries) Discussion

• Not able to use more recent data • Mortality low for elective procedures (6.7 per 1000) • Monday benchmark • Hypothesis that post-op care may have contributed to mortality • Same picture in 2 day mortality • Fall in Sunday mortality • Are junior consultants more likely to operate on a Friday? Proportions of procedures carried out by consultant experience by day of procedure in English hospitals for 2008-9 to 2010-11.

0-2 yrs 3-10 yrs ≥ 11 yrs

42.5 43.7 42.7 42.8 38.4 38.2 42.4

46.1 48.5 45.5 44.6 45 45 45.5

12 11.7 12.3 12.2 15.5 13.3 12.1

Monday Tuesday Wednesday Thursday Friday Saturday Sunday

Ruiz M, Bottle A, Aylin PP, 2015, Exploring the impact of consultants' experience on hospital mortality by day of the week: a retrospective analysis of hospital episode statistics. 2015 BMJ Qual Saf. Published Online First: doi:10.1136/bmjqs-2015-004105 Additional adjustment by consultant experience

• Adjustment makes little difference.

• Friday effect increases from OR = 1.44 to 1.48 Interpretation

• Key is higher mortality on Friday with no apparent difference in casemix • 48 hours following surgery • Frailer patients, with a high burden of co-morbidity are at higher risk of mortality at the weekend. Rapid Response

“…….The biggest increase in mortality in those operated on over Friday through to Sunday may perhaps be due to the seemingly ubiquitous practice of drastically reducing the number of doctors, in particular junior doctors, working out of hours and at weekends. A single foundation year 1 (FY1) doctor may cover a number of wards, as many as 60-100 patients. Their immediate senior, the SHO may cover twice that number or more, and so on up the ladder.”

Methods

• English patients admitted to hospital with a stroke (ICD I60-I64) from April 2009 to March 2010 • Look at 6 process and outcome indicators • Brain scan on day of admission • Thrombolysis treatment • Aspiration pneumonia • Seven-day in-hospital mortality • Discharge to usual place of residence within 56 days • Thirty day emergency readmission (all cause) • Adjusted for age, sex, socio-economic deprivation, previous emergency admissions, co-morbidities, ethnic group, source of admissions and stroke type. Results

• 93,621 stroke admissions • 9.3% died within 7 days, and 17.1% died within 30 days • 46.5% had same day brain scan • 2.6% received thrombolysis • 5.2% had aspiration pneumonia • 72.6% were discharged to their normal place of residence • 11% readmitted within 30 days Adjusted odds of stroke indicators and 95% CIs weekend versus weekday, England 2009/10

1.30

1.20

1.10

1.00

0.90 OR and 95% CI 95% and OR 0.80

0.70

0.60 Same day Thrombolysis Aspiration 7-day in-hospital Discharge to 30-day scanning pneumonia mortality usual place of emergency admission readmissions within 56 days within

Association between weekday/weekend delivery and indicators of quality and safety of care – England 1st April 2010 and 31st March 2012

1.20

1.15

1.10

1.05

1.00

0.95

0.90 Odds Ratio (95% CI) (95% Ratio Odds

0.85

0.80 Perineal tear Puerperal 3-day In-hospital Injury to Selected 3-day infection maternal perinatal neonate neonatal neonatal readmission mortality infections rradmissions Daily trends Summary

• Babies born at the weekend have an increased risk of being still born or dying in hospital within the first 7 days. • Increased complication rates at weekends, with higher rates of puerperal infection, injury to neonate and 3-day neonatal emergency readmissions • “No consistent association between outcomes and staffing was identified”

Re-analysis following comments

Indicator Adjusted OR (95% Excluding elective Adjusting for Excluding antenatal CI) – original paper caesareans induction labour stillbirths Perineal tear 1.00 - 1.00 - (0.98 to 1.03) (0.98 to 1.02) Puerperal infection 1.06 1.05 1.07 - (1.01 to 1.11) (1.00 to 1.09) (1.02 to 1.12) 3-day maternal 0.93 0.92 0.93 - readmissions (0.84 to 1.02) (0.83 to 1.02) (0.84 to 1.03) In-hospital perinatal 1.07 1.07 - 1.09 mortality (1.02 to 1.13) (1.02 to 1.12) (1.02 to 1.16) Injury to neonate 1.06 1.05 - - (1.02 to 1.09) (1.02 to 1.08) Selected neonatal 1.01 1.00 - - infections (0.98 to 1.04) (0.97 to 1.03) 3-day neonatal 1.04 1.03 - - readmissions (1.00 to 1.08) (0.99 to 1.07)

After excluding elective caesareans

Odds of death if born at the weekend compared with weekdays OR = 1.3 (1.0-1.7)

Odds of dying from anoxia at weekend compared with weekdays OR = 1.5 (1.1-2.1)

“About one in four deaths from intrapartum anoxia at term could be prevented if all women attempting vaginal birth had the same risk of this event as women delivering during the normal working week” What’s the explanation?

• Conflicting evidence? 2 Studies cited for no weekend UK studies effect (all non-UK) 1.8

1.6 4,133,346

1.4 5,271,327

14,217,640 14,818,374 1.2 93,621

1,349,599 6,923 4,317,866 49,337 294 36,734 98758 1,578 Odds/Hazard/Rate Ratio Odds/Hazard/Rate 1

0.8

Friday (Elective) Friday

Weekend (EVH) Weekend (EVH) Weekend

Weekend (Births) Weekend

Weekend (Stroke) Weekend (Stroke) Weekend (Stroke) Weekend

Weekend (Elective) Weekend

Weekend (Infection) Weekend

weekend (Emergency) weekend (Emergency) weekend (Emergency) weekend

Sunday (All admissions) (All Sunday admissions) (All Sunday

Saturday (All admissions) (All Saturday admissions) (All Saturday Aylin 2010 Freemantle 2010 Freemantle 2015 Handel Aylin 2013 (England) Palmer Palmer Mikulich Byun 2012 Kazley Kevin 2010 Myers Turin 2008 (England) (England) (England) (Scotland) 2012 2015 2010 (Korea) 2010 (US) (Canada) 2009 (US) (Japan) (Dublin) What’s the explanation?

• Conflicting evidence? • Coding? • Statistical artefact? Imperial Patient Safety Translational Research Centre “The new research based on the Oxford Vascular Study - which has been sent to the British Medical Journal but not published - comes as the BMA union starts five days of fresh talks over the junior doctors' contract”

“His team studied 1,693 stroke hospital admissions and found 638 (38%) were lower-level incidents that had been put in the wrong category.”

Rothwell et al. Administrative classification

Classification Apparent “incorrect coding” • Hospital admissions • Only admitted patients • Elective admissions • Elective admissions (293) • Subdural/extradural haemorrhage • Subdural/extradural haemorrhage (55) • Inpatient events after emergency • Inpatient events after emergency admission admission (9) • Inpatient events after elective • Inpatient events after elective admission admission (39) • Transfers from hospital • Transfers from hospital (34)

• Episodes? • Admission date wrong (24) • Spells? • GP information wrong (20) • Superspells? • Unknown (12) • Cancelled admission (15) 2 UK studies Studies cited for no 1.8 weekend effect

1.6 4,133,346

1.4 5,271,327

14,217,640 14,818,374 940,859 1.2 93,621

1,349,599 6,923 4,317,866 49,337 294 36,734 3,715,727

1 98758 1,578 Odds/Hazard/Rate Ratio Odds/Hazard/Rate 1,292 0.8

0.6

Weekend (EVH) Weekend

Friday (Elective) Friday

Weekend (Births) Weekend

Weekend (Stroke) Weekend

Weekend (via A&E) (via Weekend

Weekend (Elective) Weekend

Weekend (Infection) Weekend

weekend (Emergency) weekend

Sunday (All admissions) (All Sunday

Saturday (All admissions) (All Saturday

Saturday (All admissions) (All Saturday Weekend (Direct admissions) (Direct Weekend Aylin 2010 Freemantle 2010 Freemantle 2015 Handel Aylin 2013 (England) Palmer Palmer Mikulich Byun Kazley Kevin Myers Turin Rothwell Meacock 2016 (UK) (England) (England) (England) (Scotland) 2012 2015 2010 2012 2010 (US) 2010 2009 (US) 2008 2016 (UK) (Dublin) (Korea) (Canada) (Japan)

Meacock et al.

• Examined A&E attendances as well as admitted patients. • “There are fewer deaths following hospital admission at weekends. “ • “Higher mortality rates at weekends are found only amongst the subset of patients who are admitted.” • “Fewer and sicker patients are admitted at weekends than during the week.”

Imperial Patient Safety Translational Research Centre Meacock et al.

• Paper shows a higher mortality rate for patients admitted at the weekend • True for both admissions via A & E and direct admissions from the community. In line with previous work. • “Fewer and sicker patients are admitted at weekends than during the week. • Their own figures suggest patients admitted at weekend not sicker - younger, fewer co-morbidities etc. • A&E attendances - a questionable choice of denominator

Imperial Patient Safety Translational Research Centre What other explanations are there? Nursing and outcomes

• “An increase in a nurses' workload by one patient increased the likelihood of an inpatient dying within 30 days of admission by 7% (odds ratio 1·068, 95% CI 1·031–1·106), and every 10% increase in bachelor's degree nurses was associated with a decrease in this likelihood by 7% (0·929, 0·886–0·973).”

• “These associations imply that patients in hospitals in which 60% of nurses had bachelor's degrees and nurses cared for an average of six patients would have almost 30% lower mortality than patients in hospitals in which only 30% of nurses had bachelor's degrees and nurses cared for an average of eight patients.” Conclusions

• “Mortality outcomes after stroke are associated with the intensity of weekend staffing by registered nurses but not 7-d/wk ward rounds by stroke specialist physicians”

• “The findings have implications for quality improvement and resource allocation in stroke care”

Summary

• Lots of literature supporting the weekend effect (using variety of clinical and administrative data) • Not explained away by casemix or data quality • Clinical datasets • Relationships with other factors • Pattern not just in mortality • Differences in patterns of care • Mechanism still unclear • Failure to rescue? • Further work required on staffing levels (difficult to do), and other contributing factors.

• Interventions Imperial Patient Safety Translational Research Centre • National Cardiac Arrest Audit (NCAA)

• 27,700 in-hospital cardiac arrest

• Odds of survival weekend daytime vs. weekday daytime = 0.72 (0.68-76) Stroke mortality England 2008-2014 Acknowledgements

Dr Alex Bottle Prof Derek Bell Dr Roxana Alexandrescu Dr William Palmer Dr Min Hua Jen Dr Milagros Ruiz Prof Azeem Majeed