Insights into temporal patterns of hospital patient safety from routinely collected electronic data
The last two decades have seen an unprecedented growth in initiatives aimed to improve patient safety. For the most part, however, evidence of their impact remains controversial. At the same time, the healthcare industry has experienced an also unprecedented growth in the amount and variety of available electronic data.
In this paper, we provide a review of the use of routinely collected electronic data in the identification, analysis and surveillance of temporal patterns of patient safety.
Two important temporal patterns of the safety of hospitalised patients were identified and discussed: long-term trends related to changes in clinical practice and healthcare policy; and shorter term patterns related to variations in workforce and resources. We found that consistency in reporting is intrinsically related to availability of large-scale, fit-for-purpose data. Consistent reported trends of patient harms included an increase in the incidence of post-operative sepsis and a decrease in central-line associated bloodstream infections. Improvement in the treatment of specific diseases, such as cardiac conditions, has also been demonstrated. Linkage of hospital data with other datasets provides essential temporal information about errors, as well as information about unsuspected system deficiencies. It has played an important role in the measurement and analysis of the effects of off-hours hospital operation.
Measuring temporal patterns of patient safety is still inadequate with electronic health records not yet playing an important role. Patient safety interventions should not be implemented without a strategy for continuous monitoring of their effect.
KeywordsPatient safety Electronic health records Health services research
Administrators, healthcare providers and researchers struggle to decide which practices are most cost-effective in the improvement of patient safety. A major reason behind this problem is the fragmented and specialised nature of modern healthcare systems. This fragmentation occurs at different levels, among in-patient, out-patient and general practice care as well as among hospital wards. This results in poor information flows, disjointed care and inadequate boundaries of legal and financial responsibilities, all of which negatively influence quality of care and patient safety, and make it harder to understand and measure it [1, 2]. Furthermore, hospitals, like any other organisation, need to adapt to changes in their environment, such as ageing population, more sophisticated technologies, temporal disease patterns, variations in workforce and changes in funding . Appropriate continuous measurement of temporal patterns of patient safety in and beyond the hospital has become essential to inform hospital patient safety strategies.
With the growing availability of electronic routinely collected data inside hospitals and across the healthcare system, we have the opportunity to track patient safety over time and explore its temporal 'rhythms'. One of the criticisms of the use of routinely collected data for measuring quality of care is the fact that data may not be fit for purpose. Many measure gaps have been identified e.g. related to patient functional status, patient at the end of life, continuity of care, and some important process measures. Nevertheless, the continuous growth of the adoption of electronic health records (EHRs), new ways of sharing data via distributed data networks, together with better standardisation will slowly make the task of measuring patient safety easier.
Summary of patient safety related temporal patterns measured by routinely collected data.
Long term trends:
Effect of changes in clinical practice and healthcare policy
Temporal trends in adverse events
• Hospital administrative data (mostly using PSI)
• Registries (e.g. cardiac-related)
• National Surveillance systems (e.g. nosocomial infections)
• Pharmacy and clinical laboratory data
• Electronic Health Records
Temporal trends in performance measures
Changes in process measures and patient outcomes associated with patient-safety interventions
• Surgical safety checklist for reduction of surgical AEs.
• Subglottic secretion drainage for the prevention of ventilator-associated pneumonia .
• Sterile barriers and antibiotic-impregnated catheters for the reduction of catheter-related blood stream infections 
Hourly, weekly and monthly variations: Effect of changes in workforce and resources
Weekend and after-hours effect
• Hospital administrative data
• Registries (e.g. death, cardiac surgery)
• Increased in-hospital mortality for weekend admissions among some patient groups 77
• Increased 7 day post-admission (in-hospital and post-discharge) mortality for weekend admissions for some patient groups 
• Only 5% of six urgent procedures were performed on the weekend 
• Increase in mortality and decrease in efficiency after new residents after influx of junior residents 
• Increased LOS and mortality rates in teaching hospitals related to residency turnover 
• Significant spike in fatal medication errors during July 
Measuring hospital patient safety using routinely collected electronic data
The oldest and most widely-used measure of patient safety is mortality. Without appropriate context, however, this important patient outcome is a crude measure of safety since it does not provide a complete picture. Indeed, the use of hospital standardised mortality ratios as a measure of quality of care has been heavily criticised [4, 5]. Part of the problem stems from our inability to predict when death was preventable and, in particular, when patients are at the end of their natural life [4, 6]. Nevertheless, for specific subpopulations and in conjunction with other relevant information, mortality remains an important outcome measure of safety. Other outcomes of care that can be easily extracted from administrative datasets, and that have been correlated with patient harm are hospital length of stay  and unplanned hospital readmissions, particularly in relation to drug adverse events . Linkage of hospital administrative data with death registries, emergency department visits data, public and private claims data, and other administrative databases have proved useful to measure mortality post-discharge, re-admissions, or history of previous hospitalisations.
The patient outcome measures most closely associated to patient safety are potentially preventable adverse events, as defined by the Institute of Medicine . Traditionally, the gold standard for reporting these patient harms is through manual review of medical records by expert clinicians (see review ). This is a labour intensive job that requires at least two independent evaluators, who often present low to moderate rates of agreement. In order to facilitate this task, special methods to systematically review medical records to screen for 'triggers' of potential patient harm have been developed. The most commonly used is the Global Trigger Tool .
A more cost-effective method of detecting patient harm events can be achieved using routinely electronically captured information. Automated screening using administrative and claim records searches for adverse events among diagnosis and procedural codes. For example, the Medicare Patient Safety Monitoring System  is a US surveillance system that uses administrative and inpatient Medicare discharges to screen for adverse events. Medical records of the selected patients are then reviewed by experts, reducing the cost per chart reviewed. One drawback of this system is the potential for missing adverse events. The Agency for Healthcare Research and Quality also developed a set of Patient Safety Indicators (PSI) to recognise patient safety incidents from administrative data [13, 14]. The limitations of administrative data, in general, and of the PSIs in particular, have been pointed out in several studies . Mostly, administrative coding was generated for the purpose of reimbursement, and therefore lacks clinical content and context. Some adverse events are poorly represented by PSIs, which have been found to have positive predictive values (PPV) ranging from 28% for postoperative hip fracture to 87% for postoperative wound dehiscence . Nevertheless, the creation of newer codes (e.g. ) and, in particular, of present-on-admission (POA) flags (to discriminate between pre-existing and hospital-acquired conditions) [18, 19] have increased the validity of these measures.
Some of the limitations encountered when using administrative data alone can be reduced when combined with other electronic datasets, such as pharmacy and clinical laboratory data. These additional sources of information increase the knowledge of illness severity and can signal specific adverse events such as sudden adverse drug reactions [20, 21]. Another useful source of routinely collected information easily linked to administrative data and now also implemented in electronic format is discharge summaries. The clinical narratives of discharge summaries have proven useful in increasing the specificity of adverse event detection tools [22, 23]. Because a significant percentage of patient safety events take place post-discharge  or in an outpatient setting , linking datasets across types of care are needed to measure the full extent of the effect of hospital adverse events.
Eventually, information integrated in EHRs will provide the ideal source for automated detection of patient safety events. The use of EHRs allows better identification of clinically relevant patient groups [20, 21, 26, 27] and has the potential to greatly improve the cost-effectiveness of audit processes . Also, it provides a way to analyse the health care system as a whole and not as independent health care units. Already, automated identification of drug adverse events and hospital-acquired infections has been implemented using information technology (see e.g. ). Furthermore, in contrast to administrative and claims data, EHRs are recorded in near real time allowing for earlier surveillance of patient safety . Although levels of adoption of EHRs are still low [30, 31], and the quality of EHRs today is highly variable , as the secondary use of EHRs becomes more prevalent, and appropriate standards are introduced and validated, we expect availability and quality of EHR data to improve.
National surveillance systems and registries offer high-quality long-term counts of specific patient safety events. One of the oldest running surveillance systems of hospital adverse events is the Centers for Disease Control and Prevention's National Nosocomial Infections Surveillance (NNIS) system  (now National Healthcare Safety Network), established in 1970. Data is collected using standardised protocols and it is classified by hospital ward, associated clinical procedure or device, and infection type using information from pathology. Registries were designed to collect patient-level, fit-for-purpose, data around specific interventions or conditions in order to be able to compare institutions, devices/procedures and patients, as well as to monitor temporal trends. The most ubiquitous registries relate to cardiac surgery and cardiac interventions .
Incident reporting systems, where hospital personnel voluntarily and confidentially report incidents, represent another repository of patient safety events. Data from established large incident reporting systems usually contain a structured taxonomy that classifies the patient safety event together with free narrative describing the event . Although this data represents only a very small (under 5%) and biased sample (mainly identification, falls and medication errors) of actual patient safety events, it provides complementary information about near misses and the clinical context and contributing factors leading to patient harm [36, 37]. Studies of incidents reported by patients after hospitalisation suggest that patient reports are generally reliable [38, 39]. However, preferred ways of soliciting patient information still involve resource-consuming methods such as in-person or phone interviews. Adding questions regarding adverse events to current electronic hospital patient surveys, or extending existing incident report systems to patients are some of the cost-effective ways to build these important consumer databases. It is also possible to query and analyse electronic datasets of medical-malpractice claims. This data also represents a very small and biased fraction of all the adverse events related to medical negligence [37, 40].
Due to limitations in measuring patient harms, process of care measures have been included as additional indicators of patient safety, particularly in the context of evaluation of hospital performance . Process measures should have a demonstrable causal relation with multiple patient outcomes. Examples of evidence-based process measures are antibiotic prophylaxis before surgery  and aspirin at arrival for patients with acute myocardial infarction . Most processes of care measures are only captured in medical records or registries.
The analysis of large sets of routine clinical data plays an important role in post-marketing surveillance of harmful or ineffective treatment plans. Many adverse events associated with drugs or devices are too infrequent to be detected in randomised clinical trials, or may affect patient populations with specific co-morbidities, often excluded from experimental studies. Observational studies using large routine clinical datasets can also generate hypotheses, which can guide clinical trial design, thus providing a safer and more efficient way of generating medical evidence.
Temporal trends: the effect of changes in clinical practice and healthcare policy
Little has been said regarding the sustained widespread effects on patient safety that arise from changes in clinical practice and healthcare policy. Observational studies designed to analyse the effect of safety interventions often report short-term changes in local settings and are sometimes controversial [44, 45, 46, 47]. Reasons for this poor surveillance include the need to balance sustained improvement versus the possibility of improvement trends unrelated to the intervention, lack of good quality data that is fit for purpose, and the complexity in evaluating patient safety practice [48, 49]. On the other hand, studies analysing observed long-term, large-scale temporal trends in processes of care and patient outcomes cannot easily unravel the contribution of specific interventions; and temporal electronic data of detailed adoption and implementation of new clinical practices and policies is generally lacking. An additional difficulty arises from the need to account for changes in recording and reporting methods over time.
The temporal evolution in the incidence of hospital patient harm events has been measured using review of medical records with the Global Trigger Tool (GTT) ; PSIs [51, 52, 53, 54, 55] and diagnostic codes  from administrative datasets; disease-specific performance measures and selected outcomes collected from registries [57, 58] and national surveillance programs [59, 60, 61, 62, 63] and malpractice claims .
Reasons behind these observed changes in the rates of specific adverse events are not usually properly elucidated and discussed in the literature. It is believed that sustained changes are achieved via multistep, multilevel efforts involving clinicians, regulatory agencies and hospital administrators alike. A good example of a successful sustained reduction in patient harm is the prevention of bloodstream infections. Trends in central line-associated bloodstream infections in the US show a marked decrease since the late 90 s [62, 63, 65]. Pronovost and colleagues  describe the effort behind this success as a combination of measurement, dissemination and mobilisation across many stakeholders. Evidence on the history of the evolution of safety in hospitals also indicates that the greatest safety gains often come from the introduction and subsequent widespread adoption of safer drugs and procedures.
Analyses of performance indicators for selected conditions have found consistent improvement among a large random sample of inpatient datasets in the US in 2001 as compared to 1998 ; as well as across 3,000 accredited US hospitals in 2004 as compared to 2002 . In particular, treatment of cardiac conditions, such as acute myocardial infarction, has undergone numerous well-documented changes in the last couple of decades, which have significantly improved the safety of patients suffering from these conditions [34, 58, 59, 61, 67]. Improvement was observed in relation to the adoption of new techniques (e.g. use of coronary stent instead of atherectomy devices) and evidence-based recommendations (e.g. rapid administration of thrombolytic therapy), as well as in patient outcomes (e.g. length of stay, death).
There is no lack of studies relating specific safety interventions to improvements in the reduction of harm. A discussion on evidence-based patient safety practices can be found at . The best understood patient safety practices with higher strength of evidence relate to focus interventions aimed at reducing specific patient harms. Examples of successful interventions include the surgical safety checklist , which has been linked to a significant reduction in death and complications associated with non-cardiac surgery across the world; the use of subglottic secretion drainage for the prevention of ventilator-associated pneumonia ; and the use of sterile barriers and antibiotic-impregnated catheters for the reduction of catheter-related blood stream infections . Analyses of the impact of more general patient safety interventions are often controversial or inconclusive. For example, some of the more recent, across the board, quality of care interventions in the US: pay for performance [70, 71], restriction of specific procedures to centers of excellence [56, 72], and work hour regulations [47, 73], lack positive conclusive evidence.
Examinations of trends in patient safety require the use of large, fit-for-purpose data, such as that extracted from longitudinal medical records. Available temporal trends show a mixed picture with improvement in the treatment of some conditions and worsening or no-change in others. There have been extensive efforts on improving patient safety across the world. However, due to limitations in measures and data sources, the effects of most of these efforts remain unclear.
Hourly, weekly and monthly variations: the effect of changes in workforce and resources
When patients go to hospital, they expect to receive high levels of safety and care at all times. However, using routinely collected electronic data, several studies have demonstrated the existence of higher risk of death and adverse events at specific times of the day, days of the week, and months of the year. These periods of decreased safety have been related to both changes in staff numbers and composition, as well as lack of access to specialised clinical facilities.
It is well known that hospitals during weekdays between approximately 8 am and 6 pm and hospitals during off-hours (weekends and after-hours) are two very different clinical environments . Off-hours hospitals have only a small percentage of administrative and clinical teams, with no senior managers, almost no consultants or specialists and significantly lower nurse-to-patient ratios. These differences in staff composition and numbers are known to worry both junior doctors as well as hospital chief executive officers. Hospital residents perceive excessive work hours, inadequate supervision and problems with handoffs as the most common reasons for mistakes . Similarly, nurses perceive personal neglect, heavy workload and changes in staff as the main factors in medication errors . The relationship between staff numbers, particularly nurse staffing, and patient outcomes has also been documented (see e.g. ). On the other hand, access to some clinical facilities, such as diagnostic testing and specialised theaters, is also often not available outside normal operating hours. This reality has made researchers hypothesise a lower quality of care during weekends and after-hours, particularly for conditions that require complex immediate care outside emergency departments and intensive care units .
When looking at the 'weekend effect' more closely, researchers have found that there are measurable differences in patient treatments. In particular, it has been observed that the incidence and timing of important interventional procedures differs for weekend vs. weekday patients. For example, off-hours patients presenting with acute myocardial infarction are less likely to receive immediate cardiac procedures [81, 83]; and experience substantially longer door-to-balloon times . Indeed, Kostis et al  found that statistically significant differences in 30 days mortality became nonsignificant after adjusting for invasive cardiac procedures. Similarly Bell et al  analysed time to six prespecified procedures among more than 100,000 emergency admissions in Canada, and found that only 5% of urgent procedures were performed on the weekend. Delays in interventional procedures have been independently related to infection complications, length of stay and mortality [90, 91].
The effect of weekends and evenings has also been found when comparing the outcomes from elective surgery among days of the week [92, 93]; for patients suffering from an in-hospital cardiac arrest and for adults admitted to and discharged from intensive care .
Another phenomenon observed using routinely collected data is the relationship between patient outcomes and the cyclic rotation of trainee doctors, which takes place in teaching hospitals. In the Northern Hemisphere the influx of new or junior residents occurs in July, and therefore the existence of this period of observed decreased patient safety has been coined the 'July effect'. Researchers hypothesise that this staff turnover disrupts established doctor teams abruptly decreasing the average experience of the workforce. Young and colleagues performed a review of the literature regarding this phenomenon in 2011 . They concluded that, in general, existing studies demonstrate an increase in mortality and a decrease in efficiency (reported as length of stay, duration of procedure and hospital charges) related to the end of the year changeovers. The 'July effect', like the 'weekend effect', does not affect all patient groups and a substantial degree of heterogeneity among studies has been found; however, its evidence is generally weaker. One of the largest studies of the effect of cohort turnover involved roughly 20% of all US hospitals for a period of 5 years . By comparing temporal trends of teaching vs non-teaching hospitals, the authors found increased average length of stay and mortality rates related to residency turnover. Another large study including over 60 million death certificates and more than 200,000 medication errors, found a significant spike in fatal medication errors at medical institutions during July .
Evidence of the impact of differences in standard of care, such as out-of-hours care, could not have been revealed without the use of large routinely collected electronic data. More informative data and further research are needed to unravel the causes behind these periods of decreased safety and provide appropriate solutions.
In this paper we provide a review of the current use of routinely collected electronic data to identify and analyse temporal patterns of hospital patient safety. We found that the science of measuring patient safety is still not properly advanced and surrounded by controversy. The gold standard of counting patient harms is still manual review of medical records, and use of full EHRs for this purpose is far from widespread. This is an important problem, since it is well known that trust in measures and the measurement process facilitate the use of the recorded information for training and improvement. For those situations in which there is suitable large-scale temporal data, it is much easier to find consistent trends. One example is the observed decrease in central line-associated bloodstream infections in the US, captured by the national surveillance program on nosocomial infections.
Assessing the impact of patient safety interventions is currently poorly done. Most studies only report pre and post intervention effects in local settings without any further monitoring of sustained change. Often these reports are controversial. We propose that interventions should always be implemented together with a strategy for continuous monitoring of the target patient outcome.
Data linkage has proved useful in the unravelling of important temporal aspects of patient safety; for example, death and birth registries, data containing the day and time of interventional procedures and clinical datasets that help to better determine the illness severity of patients. However, some useful patient-safety related datasets, such as information about staffing numbers in each hospital shift and information about locum personnel, are mostly unavailable. A notable exception is the case of some hospitals in Massachusetts which provide voluntary unit-by-unit reports on caregiver staffing levels updated on an annual basis for over 750 hospital units including emergency departments .
Some of the temporal patterns discussed in this review warrant further investigation. A worrisome increase in the incidence of post-operative sepsis, for example, has been found in several studies. New evidence on the negative effect of weekend and after-hours hospital operation has made researchers argue about what should be the most cost-effective way of improving off-hours hospital care . If the cause of the 'weekend effect' is understaffing then the strategy should be to regulate workforce. If, however, the cause is inappropriate treatment (such as delays to surgery) not related to staffing numbers then a more cost-effective strategy could be rewarding good off-hours performance. Information technology can also help reducing the gap between the 'normal' and the 'off-hours' hospital by creating safety nets and providing information for junior doctors when senior staff are not present.
Analyses of routinely collected electronic records provide essential insights into temporal patterns of hospital patient safety. They reveal the existence of non-clinical temporal patterns related to hospital workflow and allow for the monitoring of the effect of patient safety interventions. However, there are still important information gaps with respect to both patients (clinical and social status) as well as hospitals (staffing numbers, availability of specialised facilities). The use of integrated EHRs will be important to fill in some of these gaps and provide more meaningful measures of safety.
Funding and ethics statement
This work was funded by National Health and Medical Research Council (NHMRC) Program Grant 568612, and Project Grant 1045548; and approved by the NSW Population and Health Services Research Ethics Committee and the UNSW Human Research Ethics Committee. Its contents are the responsibility of the authors and their institutions and do not reflect the views of the NHMRC.
Publication costs were funded by NHMRC Program Grant 568612
This article has been published as part of Health Information Science and Systems Volume 3 Supplement 1, 2015: Proceedings of the Health Informatics Society of Australia Big Data Conference (HISA 2013). The full contents of the supplement are available online at http://www.hissjournal.com/supplements/3/S1/
- 1.Wachter RM: The end of the beginning: patient safety five years after "To Err Is Human.". Health Affairs. 2004, 23 (11): 534-545.Google Scholar
- 3.Stephen W, Nigel E, Bernd R: Investing in hospitals of the future. 2009, World Health OrganizationGoogle Scholar
- 5.Mohammed MA, Deeks JJ, Girling A, Rudge G, Carmalt M, Stevens AJ, Lilford RJ: Evidence of methodological bias in hospital standardised mortality ratios: retrospective database study of English hospitals. BMJ: British Medical Journal. 2009, 338-Google Scholar
- 9.Kohn L, Corrigan J, Donaldson M: To Err is Human. Building a Safer Health System. Committee on Quality of Health Care in America. 1999, Washington, DC: Institute of Medicine. Academy PressGoogle Scholar
- 11.Classen DC, Resar R, Griffin F, Federico F, Frankel T, Kimmel N, Whittington JC, Frankel A, Seger A, James BC: 'Global trigger tool'shows that adverse events in hospitals may be ten times greater than previously measured. Health Affairs. 2011, 30 (4): 581-589. 10.1377/hlthaff.2011.0190.CrossRefPubMedGoogle Scholar
- 12.Hunt DR, Verzier N, Abend SL, Lyder C, Jaser LJ, Safer N, Davern P: Fundamentals of Medicare patient safety surveillance: intent, relevance, and transparency. DTIC Document. 2005Google Scholar
- 14.McDonald KM, Romano PS, Geppert J, Davies SM, Duncan BW, Shojania KG, Hansen A: Measures of patient safety based on hospital administrative data-the patient safety indicators. 2002Google Scholar
- 16.Rosen AK, Itani KM, Cevasco M, Kaafarani HM, Hanchate A, Shin M, Shwartz M, Loveland S, Chen Q, Borzecki A: Validating the patient safety indicators in the Veterans Health Administration: do they accurately identify true safety events?. Medical care. 2012, 50 (1): 74-85. 10.1097/MLR.0b013e3182293edf.CrossRefPubMedGoogle Scholar
- 17.Sadeghi B, White RH, Maynard G, Zrelak P, Strater A, Hensley L, Cerese J, Romano P: Improved Coding of Postoperative Deep Vein Thrombosis and Pulmonary Embolism in Administrative Data (AHRQ Patient Safety Indicator 12) After Introduction of New ICD-9-CM Diagnosis Codes. Medical care. 2013Google Scholar
- 18.Glance LG, Osler TM, Mukamel DB, Dick AW: Impact of the present-on-admission indicator on hospital quality measurement: experience with the Agency for Healthcare Research and Quality (AHRQ) Inpatient Quality Indicators. Medical care. 2008, 46 (2): 112-119. 10.1097/MLR.0b013e318158aed6.CrossRefPubMedGoogle Scholar
- 21.Pine M, Sonneborn M, Schindler J, Stanek M, Maeda JL, Hanlon C: Harnessing the power of enhanced data for healthcare quality improvement: lessons from a Minnesota Hospital Association Pilot Project. Journal of healthcare management/American College of Healthcare Executives. 2011, 57 (6): 406-418. discussion 419-420Google Scholar
- 26.Jollis JG, Ancukiewicz M, DeLong ER, Pryor DB, Muhlbaier LH, Mark DB: Discordance of databases designed for claims payment versus clinical information systems: implications for outcomes research. Annals of Internal Medicine. 1993, 119 (8): 844-850. 10.7326/0003-4819-119-8-199310150-00011.CrossRefPubMedGoogle Scholar
- 27.Tang PC, Ralston M, Arrigotti MF, Qureshi L, Graham J: Comparison of methodologies for calculating quality measures based on administrative data versus clinical data from an electronic health record system: implications for performance measures. Journal of the American Medical Informatics Association. 2007, 14 (1): 10-15. 10.1197/jamia.M2198.PubMedCentralCrossRefPubMedGoogle Scholar
- 28.Freeman R, Moore L, García Álvarez L, Charlett A, Holmes A: Advances in electronic surveillance for healthcare-associated infections in the 21st Century: a systematic review. Journal of Hospital Infection. 2013Google Scholar
- 33.National Healthcare Safety Network (NHSN) Annual Reports. [http://www.cdc.gov/nhsn/datastat.html]
- 34.Hannan EL, Cozzens K, King III SB, Walford G, Shah NR: The New York State Cardiac Registries. Journal of the American College of Cardiology. 2012, 59 (25):Google Scholar
- 37.Levtzion-Korach O, Frankel A, Alcalai H, Keohane C, Orav J, Graydon-Baker E, Barnes J, Gordon K, Puopulo AL, Tomov EI: Integrating incident data from five reporting systems to assess patient safety: making sense of the elephant. Joint Commission Journal on Quality and Patient Safety. 2010, 36 (9): 402-410.PubMedGoogle Scholar
- 38.Weingart SN, Pagovich O, Sands DZ, Li JM, Aronson MD, Davis RB, Bates DW, Phillips RS: What Can Hospitalized Patients Tell Us About Adverse Events? Learning from Patient-Reported Incidents. Journal of General Internal Medicine. 2005, 20 (9): 830-836. 10.1111/j.1525-1497.2005.0180.x.PubMedCentralCrossRefPubMedGoogle Scholar
- 39.Weissman JS, Schneider EC, Weingart SN, Epstein AM, David-Kasdan J, Feibelmann S, Annas CL, Ridley N, Kirle L, Gatsonis C: Comparing patient-reported hospital adverse events with medical record review: do patients know something that hospitals do not?. Annals of Internal Medicine. 2008, 149 (2): 100-108. 10.7326/0003-4819-149-2-200807150-00006.CrossRefPubMedGoogle Scholar
- 40.Localio AR, Lawthers AG, Brennan TA, Laird NM, Hebert LE, Peterson LM, Newhouse JP, Weiler PC, Hiatt HH: Relation between malpractice claims and adverse events due to negligence: results of the Harvard Medical Practice Study III. New England Journal of Medicine. 1991, 325 (4): 245-251. 10.1056/NEJM199107253250405.CrossRefPubMedGoogle Scholar
- 42.Hawn MT, Itani KM, Gray SH, Vick CC, Henderson W, Houston TK: Association of timely administration of prophylactic antibiotics for major surgical procedures and surgical site infection. Journal of the American College of Surgeons. 2008, 206 (5): 814-819. 10.1016/j.jamcollsurg.2007.12.013.CrossRefPubMedGoogle Scholar
- 52.Rosen AK, Zhao S, Rivard P, Loveland S, Montez-Rath ME, Elixhauser A, Romano PS: Tracking rates of Patient Safety Indicators over time: lessons from the Veterans Administration. Medical care. 2006, 850-861.Google Scholar
- 56.Dimick JB, Nicholas LH, Ryan AM, Thumma JR, Birkmeyer JD: Bariatric Surgery Complications Before vs After Implementation of a National Policy Restricting Coverage to Centers of ExcellenceComplications Following Bariatric Procedures. JAMA. 2013, 309 (8): 792-799. 10.1001/jama.2013.755.PubMedCentralCrossRefPubMedGoogle Scholar
- 57.Anderson FA, Hirsh J, White K, Fitzgerald RH: Temporal Trends in Prevention of Venous Thromboembolism Following Primary Total Hip or Knee Arthroplasty 1996-2001 Findings From the Hip and Knee Registry. CHEST Journal. 2003, 124 (6_suppl): 349S-356S. 10.1378/chest.124.6_suppl.349S.CrossRefGoogle Scholar
- 58.Boltuch J, Sabeti S, Amighi J, Dick P, Mlekusch W, Schlager O, Ahmadi A, Minar E, Schillinger M: Procedure-related complications and early neurological adverse events of unprotected and protected carotid stenting: temporal trends in a consecutive patient series. Journal of Endovascular Therapy. 2005, 12 (5): 538-547. 10.1583/05-1615R.1.CrossRefPubMedGoogle Scholar
- 59.Rogers WJ, Canto JG, Lambrew CT, Tiefenbrunn AJ, Kinkaid B, Shoultz DA, Frederick PD, Every N: Temporal trends in the treatment of over 1.5 million patients with myocardial infarction in the U.S. from 1990 through 1999 The National Registry of Myocardial Infarction 1, 2 and 3. Journal of the American College of Cardiology. 2000, 36 (7): 2056-2063. 10.1016/S0735-1097(00)00996-7.CrossRefPubMedGoogle Scholar
- 63.Burton DC, Edwards JR, Horan TC, Jernigan JA, Fridkin SK: Methicillin-resistant Staphylococcus aureus central line-associated bloodstream infections in US intensive care units, 1997-2007. JAMA: the journal of the American Medical Association. 2009, 301 (7): 727-736. 10.1001/jama.2009.153.CrossRefPubMedGoogle Scholar
- 65.Berenholtz SM, Pronovost PJ, Lipsett PA, Hobson D, Earsing K, Farley JE, Milanovich S, Garrett-Mayer E, Winters BD, Rubin HR: Eliminating catheter-related bloodstream infections in the intensive care unit*. Critical care medicine. 2004, 32 (10): 2014-2020. 10.1097/01.CCM.0000142399.70913.2F.CrossRefPubMedGoogle Scholar
- 67.Peterson ED, Lansky AJ, Anstrom KJ, Muhlbaier LH, Popma JJ, Satler LF, Lanzilotta MJ: Evolving trends in interventional device use and outcomes: results from the National Cardiovascular Network Database. American Heart Journal. 2000, 139 (2): 198-207. 10.1016/S0002-8703(00)90226-6.CrossRefPubMedGoogle Scholar
- 68.Van Klei W, Hoff R, Van Aarnhem E, Simmermacher R, Regli L, Kappen T, van Wolfswinkel L, Kalkman C, Buhre W, Peelen L: Effects of the introduction of the WHO "Surgical Safety Checklist" on in-hospital mortality: a cohort study. Annals of surgery. 2012, 255 (1): 44-49. 10.1097/SLA.0b013e31823779ae.CrossRefPubMedGoogle Scholar
- 71.Glickman SW, Ou F-S, DeLong ER, Roe MT, Lytle BL, Mulgund J, Rumsfeld JS, Gibler WB, Ohman EM, Schulman KA: Pay for performance, quality of care, and outcomes in acute myocardial infarction. JAMA: the journal of the American Medical Association. 2007, 297 (21): 2373-2380. 10.1001/jama.297.21.2373.CrossRefPubMedGoogle Scholar
- 80.Perez-Concha O, Gallego B, Hillman K, Delaney G, Coiera E: Mortality Patterns vary across different patient groups following weekend admission: Reduced quality of care or a different patient cohort mix?. BMJ Quality & Safety. 2013, doi: 10.1136/bmjqs-2013-002218Google Scholar
- 85.Pasupathy D, Wood AM, Pell JP, Fleming M, Smith GC: Time of birth and risk of neonatal death at term: retrospective cohort study. BMJ: British Medical Journal. 2010, 341-Google Scholar
- 89.Magid DJ, Wang Y, Herrin J, McNamara RL, Bradley EH, Curtis JP, Pollack CV, French WJ, Blaney ME, Krumholz HM: Relationship between time of day, day of week, timeliness of reperfusion, and in-hospital mortality for patients with acute ST-segment elevation myocardial infarction. JAMA: the journal of the American Medical Association. 2005, 294 (7): 803-10.1001/jama.294.7.803.CrossRefPubMedGoogle Scholar
- 91.Moja L, Piatti A, Pecoraro V, Ricci C, Virgili G, Salanti G, Germagnoli L, Liberati A, Banfi G: Timing matters in hip fracture surgery: patients operated within 48 hours have better outcomes. A meta-analysis and meta-regression of over 190,000 patients. PloS one. 2012, 7 (10): e46175-10.1371/journal.pone.0046175.PubMedCentralCrossRefPubMedGoogle Scholar
- 92.Aylin P, Alexandrescu R, Jen MH, Mayer EK, Bottle A: Day of week of procedure and 30 day mortality for elective surgery: retrospective analysis of hospital episode statistics. BMJ. 2013, 346-Google Scholar
- 94.Peberdy MA, Ornato JP, Larkin GL, Braithwaite RS, Kashner TM, Carey SM, Meaney PA, Cen L, Nadkarni VM, Praestgaard AH: Survival from in-hospital cardiac arrest during nights and weekends. JAMA: the journal of the American Medical Association. 2008, 299 (7): 785-792. 10.1001/jama.299.7.785.CrossRefPubMedGoogle Scholar
- 97.Huckman RS, Barro J: Cohort turnover and productivity: the July phenomenon in teaching hospitals. National Bureau of Economic Research. 2005Google Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.