Healthcare Consulting Services
Healthcare Consulting Services with a Focus on Patient Safety Solutions and Quality Improvement Across the Health Care Continuum. Your Patient Safety Resource Solution.
November 18, 2014
Handwashing Fades at End of Shift
?Smartwatch to the Rescue
We’ve done a lot of columns on procedures done less well at the end of the day or end of the shift. Now a new study shows that compliance with handwashing also fades late in the day or toward the end of a shift. Dai and colleagues (Dai 2014) monitored handwashing by using RFID technology to determine how often and how soon healthcare workers washed their hands after entering or leaving patient rooms. Analyzing data over a 3-year period, they found that handwashing rates dropped 8.7% from the beginning to the end of a 12-hour shift. The decline was also magnified by increased work intensity and increased as individuals accumulated more total work hours the preceding week.
So are there solutions to the handwashing problem? Intermountain Healthcare may have a technological solution to this problem: the smartwatch (Terry 2014). The watch has a color-based alarm that triggers as the healthcare worker changes rooms. If the worker moves to another room it changes from green to either red or yellow to alert the worker he/she needs to wash his/her hands. Use of the watch has been associated with a reduction in infections. Commentaries from representatives of The Advisory Board and the Institute for Healthcare Improvement note that there are multiple other technological devices doing similar alerts but that there still is a need for accountability. The smartwatch and other technologies also send data back to managers so not only are there real-time reminders to wash hands but managers and medical directors can use the aggregate data for discussion with each individual healthcare worker.
Poor performance later in the day has now been noted for many procedures. In October 2014 we noted problems with laparoscopic cholecystectomies done after hours (“What Time of Day Do You Want Your Surgery?”) and in September 2009 we noted many orthopedic procedures that were problematic after-hours (September 2009 “After-Hours Surgery – Is There a Downside?”). In an upcoming column (“Another Procedure to Avoid Late in the Day or on Weekends”) we note that implantable cardioverter-defibrillator (ICD) recipients implanted in the afternoon/evening and on weekends or holidays more often experienced adverse events. We’ve also previously noted that the rate of incomplete colonoscopies increases late in the day and fewer adeomas are detected with colonoscopies later in the day (see our May 3, 2011 Patient Safety Tip of the Week “It’s All in the Timing”). And in our July 31, 2012 Patient Safety Tip of the Week “Surgical Case Duration and Miscommunications” we noted in a study of simulated surgery (Feuerbacher 2012) that residents made more errors when distracted or interrupted but all the errors occurred after 1 PM. A similar phenomenon has been reported in radiology. One study (Krupinski 2010) showed a significant reduction in diagnostic accuracy of radiologists after a day of clinical reading (average 8 hours), as measured by reduced ability to detect fractures.
So should it really come as a surprise to you that handwashing rates fall off later in the shift?
Is it fatigue that is responsible for deterioration in performance later in the day? That’s the most obvious factor. Fatigue clearly impairs cognitive processes and may lead to errors (see all our previous columns on the impact of fatigue listed at the end of today’s column). But there are other issues that might be involved. In fact, in the study by Feuerbacher (Feuerbacher 2012) fatigue was measured by several parameters and did not appear to be the primary issue.
Trying to get everything done in a condensed time period is probably one of the biggest contributing factors. When tasks are compressed into a shortened time period we have to prioritize them. While we like to think we always prioritize the most important ones first, that’s not always the case. In our November 26, 2013 Patient Safety Tip of the Week “Missed Care Opportunities” we noted that tasks which have the most immediate consequences tend to get top priority and ones that have delayed consequences are often deferred or omitted. Handwashing is an activity in which the adverse consequences (infection, etc.) are typically delayed so it’s not surprising it may get lower priorities.
Think about yourself near the end of a day or shift at work. You may be quite alert and not fatigued but lots of other things begin to pop into your consciousness and compete with your tasks at hand. You may be thinking about all the things you still need to do today before you leave. Or getting ready for tomorrow morning’s meeting. Or what you are going to do once you get home tonight. Or packing for your weekend trip.
Hunger also can be distracting. Watch your audience the next time you give an 11AM lecture!
And though it’s unlikely to contribute to the handwashing issue, monotony may contribute to some of the late-in-the-day or end-of-shift errors. In our May 3, 2011 Patient Safety Tip of the Week “It’s All in the Timing” we noted that fewer abnormalities are found by pathologists or cytology techs looking at slides for long periods (hence the interest in automated procedures to screen specimens for abnormalities). Nurses or technicians monitoring telemetry screens are also less likely to detect abnormalities when watching monitors for long periods. Errors related to monotony have been seen in other industries such as trucking, banking, inspecting goods, measuring parts, lifeguard surveillance, railway transportation, etc.
So what are the take-home lessons from all this? First, there is obvious utility in looking at various outcome measures not just in the aggregate but also by time of day (and maybe day of the week as well). Developing flexible scheduling for surgeries and procedures may reduce the late-in-the-day add-on cases. Limiting the workload in some circumstances may make sense. The colonoscopy issue was addressed by imposing a cap of 3-hours per session for individual colonoscopists. Breaking up routines can address the issue of monotony in a whole variety of processes.
Unfortunately, none of these really help the handwashing issue. Maybe that smartwatch really is the one thing that will make me take 15 seconds to wash my hands. Uh-oh! That smart sink the hospital installed just let me know that 15 seconds is not long enough for handwashing!
Some of our previous columns on the “weekend” and “after hours” effects:
Some of our other columns on the role of fatigue in Patient Safety:
Dai H, Milkman KL, Hoffman DA, Staats, BR. The Impact of Time at Work and Time Off From Work on Rule Compliance: The Case of Hand Hygiene in Health Care. Journal of Applied Psychology 2014; Published online Nov. 3, 2014
Terry K. New Smart Watch May Improve Hand Hygiene in Hospitals. Medscape 2014; Nov 10, 2014
Feuerbacher RL, Funk KH, Spight DH, et al. Realistic Distractions and Interruptions That Impair Simulated Surgical Performance by Novice Surgeons. Arch Surg 2012; 147(11): 1026-1030 published online first July 2012
Krupinski EA, Berbaum KS, Caldwell RT, et al. Long Radiology Workdays Reduce Detection and Accommodation Accuracy. Journal of the American College of Radiology 2010; 7(9): 698-704
To get "Patient Safety Tip of the Week " emailed to you, click here and enter "subscribe" in the subject field.
If you don't see the search term you expected to see here, its probably because that tip already went to our Tip of the Week Archive. We do a new tip every week. Click here to search the entire site or you can Go to Tip of the Week Archive a patient safety resource solution loaded with tips, tools, and techniques you can use in your patient safety and quality improvement initiatives. Or it may have moved to our What's New Archive.
Click here to see the consulting services and patient safety solutions that we provide.
For a long time we’ve felt that performance measures for VTE prophylaxis may be resulting in many low-risk patients receiving unnecessary prophylaxis. Now another study has questioned whether we are overusing prophylaxis in general medical patients. Flanders et al reported the results of a large multihospital performance improvement collaborative in Michigan (Flanders 2014). They looked at VTE events occurring within 90 days of hospital admission and stratified the results by hospital performance on a VTE prophylaxis measure. Overall, the rate of VTE events in this medical population was very low. When stratified into tertiles (prophylaxis rates 85.8%, 72.6% and 55.5% for high, moderate and low hospitals respectively) they found no difference in the VTE rates. Their results suggest that efforts to increase the rates of VTE prophylaxis in non-critically ill general medical patients may not substantially reduce the rate of VTE.
The accompanying commentary (Rothberg 2014) is also very thoughtful. One interpretation is that giving prophylaxis to large numbers of low-risk medical patients does not significantly reduce VTE and exposes patients unnecessarily to painful injections, creates excess costs and might increase the risk of bleeding. Rothberg notes that the fundamental problem is lack of validated tools for risk assessment in this population. The Caprini model has been validated in surgical, but not medical, patients. Rothberg suggests that the Padua Prediction Score (recommended in the most recent ACCP guidelines for VTE prophylaxis) might prove to be a better tool in medical patients but still needs to be validated in a medical population in the US because of differences between patients in the Italy and the US.
Our April 2014 What’s New in the Patient Safety World column “Another Rap on the VTE Prophylaxis Measure” discussed another study that questioned the utility of the VTE prophylaxis measure (JohnBull 2014). Those authors used publicly reported data from CMS to see if there was a correlation between VTE prophylaxis rates and outcome rates and found no correlation. They found that VTE rates at hospitals reporting 100% compliance with VTE prophylaxis were no different than those at hospitals in the bottom quintile of prophylaxis rates.
In our February 15, 2011 Patient Safety Tip of the Week “Controversies in VTE Prophylaxis” we highlighted a study by some very respected investigators in the surgical quality improvement field (Qadan 2011) that questioned the current recommendations on venous thrombembolism (VTE) prophylaxis in elective major surgery. The authors collected data on DVT and PE in patients electively undergoing 4 major surgical procedures (colorectal resection, total knee replacement, total hip replacement, and hysterectomy) from a large database from a consortium of academic medical centers for two periods of time (2003-2004 vs. 2007-2008). The study demonstrated a substantial increase in the use of pharmacologic DVT prophylaxis between the two time periods. Yet the rates of DVT and PE were not significantly impacted by this increased use of such prophylaxis. Moreover, the overall rates of DVT and PE were actually quite low and the rates in patients who did not receive pharmacoprophylaxis actually decreased between the two time periods. The authors concluded that this may show that clinical judgment of physicians in choosing which patients need pharmacoprophylaxis is remarkable.
In our November 2013 What’s New in the Patient Safety World column “Are VTE Measures Flawed as Quality Measures?” we highlighted a study (Bilimoria 2013) that looked at almost a million surgical patients in almost 3000 hospitals for VTE prophylaxis rates, VTE event rates, and use of imaging in VTE screening. They found that greater hospital VTE prophylaxis adherence rates were weakly associated with worse risk-adjusted VTE event rates. When they looked at hospitals with higher “structural” quality scores (based on 8 factors commonly thought to be associated with higher commitment to quality) they found higher VTE prophylaxis adherence rates but worse risk-adjusted VTE rates. Risk-adjusted VTE rates increased significantly with VTE imaging use rates in a stepwise fashion, leading to their conclusion that surveillance bias limits the usefulness of the VTE quality measure for hospitals.
The time has come to refine the VTE prophylaxis performance measures to ensure that we are not creating unintended consequences. It would not be the first time performance measures have created such. We all recall the original CMS measure for prompt antibiotic administration in patients with community acquired pneumonia resulted in many patients not having pneumonia at all receiving unnecessary antibiotics.
But critical to refining the VTE prophylaxis measure, if it is to be kept at all, will be to standardize on a validated risk assessment tool and remove the surveillance bias that may affect rates.
Some of our prior columns on issues related to VTE prophylaxis measures:
February 15, 2011 “Controversies in VTE Prophylaxis”
November 2013 “Are VTE Measures Flawed as Quality Measures?”
April 2014 “Another Rap on the VTE Prophylaxis Measure”
Flanders SA, Greene T, Grant P, et al. Hospital Performance for Pharmacologic Venous Thromboembolism Prophylaxis and Rate of Venous Thromboembolism. A Cohort Study. JAMA Intern Med. 2014; 174(10): 1577-1584
Rothberg MB. Venous Thromboembolism Prophylaxis for Medical PatientsWho Needs It? JAMA Intern Med 2014; 174(10): 1585-1586
Qadan M, Polk HC, Hohmann SF, Fry DE. A reassessment of needs and practice patterns in pharmacologic prophylaxis of venous thromboembolism following elective major surgery. Ann Surg 2011; 253(2): 215-220
JohnBull EA, Lau BD, Schneider EB, et al. No Association Between Hospital-Reported Perioperative Venous Thromboembolism Prophylaxis and Outcome Rates in Publicly Reported Data (Research Letter). JAMA Surgery 2014; online first February 5, 2014
Bilimoria KY, Chung J, Ju MH, et al. Evaluation of Surveillance Bias and the Validity of the Venous Thromboembolism Quality Measure. JAMA 2013; 310(14): 1482-1489
A new study investigated out-of-hospital medication errors involving children (Smith 2014). The investigators used a database of reports from poison control centers nationwide. Between 2002 and 2012 they found 63,000 such errors annually in children under the age of 6 or one error every 8 minutes in the US. They acknowledge this is likely an underestimate since the database captures only those cases that were reported to poison control centers.
Most of the incidents (96.9%) occurred in the child’s own home but some occurred in other residences or in school. Fortunately, most did not cause serious harm to children and most did not require management or monitoring in a healthcare facility. Analgesics, cough and cold preparations, antihistamines and antibiotics accounted for the vast majority of incidents. Boys accounted for a slightly higher percentage of cases and there was a seasonal pattern with higher incidence in winter (more prescriptions for offending medications occur during winter months). The percentage due to cough and cold preparations decreased considerably over the study period, largely due to efforts to reduce their inappropriate prescription.
Being inadvertently given a medication twice accounted for 27% of medication errors, followed by other incorrect dose (17.8%), confused units of measure (8.2%), and wrong medication (7.8%).
The incidence of medication errors increased in younger children, with more than 25% of cases occurring in children under the age of 1 year. Liquid medications were more often involved in younger patients.
The authors note several opportunities to reduce the occurrence of such medication errors in children. Efforts at educating parents on correct use of dosing devices and better instructions on labeling and packaging are important. Moreover, adherence to evidence-based guidelines for prescription of drugs in the analgesic, antihistamine, and antimicrobial categories should be emphasized.
Language barriers, of course, may play a role in pediatric medication errors. So may health literacy issues. For years when we have talked about health literacy we have focused on reading levels and reading comprehension. But in our June 2012 What’s New in the Patient Safety World column “Parents' Math Ability Matters” we noted a study (AAP 2012) that showed that parents’ mathematics skills, independent of reading skills, may play a big role in some pediatric medication errors. The study was done by Marrese et al. and presented as an abstract at the Pediatric Academic Societies (PAS) annual meeting April 28, 2012 (Marrese 2012). It showed parents with math skills at the third grade level or below were five times more likely to measure the wrong dose of medication for their child than those with skills at the sixth grade level or higher. While about a third of the parents had low reading skills, 83% had poor “numeracy” skills, with 27% having skills at or below the third grade level. Parents with low numeracy may especially prone to make errors in tasks requiring dose measurement or measurement conversions.
The study highlights the need to address numeracy skills of parents when communicating medication instructions (we suspect the same is likely to apply to adult medication errors as well). They also provide as an example having providers review and give parents pictures of dosing instruments filled to the correct amount for that prescription.
And a study earlier this year showed that parent’s measurement and dosing errors are common (Yin 2014). 39.4% of parents made an error in measurement of the intended dose and 41.1% made an error in the prescribed dose. Furthermore, 16.7% used a nonstandard instrument. Compared with parents who used milliliter-only, parents who used teaspoon or tablespoon units had twice the odds of making an error with the intended and prescribed dose. Associations were greater for parents with low health literacy and non–English speakers. Nonstandard instrument use partially mediated teaspoon and tablespoon–associated measurement errors. The authors conclude that their findings support a milliliter-only standard to reduce medication errors.
We also refer you to our May 7, 2013 Patient Safety Tip of the Week “Drug Errors in the Home” which had many observations and recommendations regarding pediatric patients in the home as well as adult patients.
Some of our other columns on pediatric medication errors:
November 2007 “1000-fold Overdoses by Transposing mg for micrograms”
December 2007 “1000-fold Heparin Overdoses Back in the News Again”
September 9, 2008 “Less is More and Do You Really Need that Decimal?”
June 28, 2011 “Long-Acting and Extended-Release Opioid Dangers”
September 13, 2011 “Do You Use Fentanyl Transdermal Patches Safely?”
September 2011 “Dose Rounding in Pediatrics”
April 17, 2012 “10x Dose Errors in Pediatrics”
May 2012 “Another Fentanyl Patch Warning from FDA”
June 2012 “Parents’ Math Ability Matters”
Septembrer 2012 “FDA Warning on Codeine Use in Children Following Tonsillectomy”
May 7, 2013 “Drug Errors in the Home”
May 2014 “Pediatric Codeine Prescriptions in the ER”
Smith MD, Spiller HA, Casavant MJ, et al. Out-of-Hospital Medication Errors Among Young Children in the United States, 2002-2012. Pediatrics 2014; 134: 867–876 published online October 20, 2014
American Academy of Pediatrics (AAP). Parents’ Poor Math Skills May Lead to Medication Errors. AAP press release April 28, 2012
Marrese C, Dreyer B, Mendelsohn A, Moreira H, Yin HS. Parent Medication Dosing Errors: Role of Health Literacy and Numeracy (abstract). Pediatric Academic Societies (PAS) annual meeting April 28, 2012
Yin HS, Dreyer BP, Ugboaja DC, et al. Unit of Measurement Used and Parent Medication Dosing Errors. Pediatrics 2014; 134(2): e354-e361; published ahead of print July 14, 2014
In our July 1, 2014 Patient Safety Tip of the Week “Interruptions and Radiologists” we took a look at the impact of interruptions and distractions on the workflow of radiologists. A study (Yu 2014) found that during a typical 8PM to 8AM overnight shift there was an average of 72 telephone calls, with a median call duration 57 seconds, and the average time spent on the phone was 108 minutes. The median interval from the start of one telephone call to the start of the next ranged from 3 to 10 minutes, depending on the time of day. There was also a correlation between volume of phone calls and the volume of CT scans being done (volume of other imaging studies was not measured as part of this study). That study did not include any measure of image interpretation accuracy or disparities between interpretations by the on-call radiologist and any subsequent interpretations.
But now another study has looked at the impact of telephone calls on radiology residents on-call to determine whether there was a relationship between these and discrepancies on reports (Balint 2014). While there was a only a slight difference in total phone calls per shift between those shifts with and without report discrepancies, there was a statistically significant increase in the average number of phone calls in the 1 hour preceding the generation of a discrepant preliminary report (4.23 vs. 3.24 calls). The authors suggest that one additional phone call during the hour preceding the generation of a discrepant preliminary report resulted in a 12% increased likelihood of a resident error.
One of the strategies they recommend to prevent such interruptions is to have other staff handle phone calls. Additonal potential strategies include interruption-free zones and having a separate radiologist or radiology resident handle consultations. The previous study by Yu and colleagues noted that posting preliminary reports on the electronic medical record has likely had a beneficial effect on frequency of calls. They, too, have also begun having medical students assist the on-call radiologist by answering the phone and triaging imaging reports. We added that radiology physician assistants can help with things like contrast injections, etc. during high activity periods that might also interrupt radiologists’ reading. Hospitals having the luxury of larger radiology staffs might have a dedicated second radiologist during high volume periods whose sole responsibility is interpreting images. Note that the latter might also be reading images off-site via teleradiology.
We also stressed that much time can be wasted in tracking down the appropriate physician when communicating significant findings. So anything you can do to facilitate identification of the responsible physician would be a positive step.
Prior Patient Safety Tips of the Week dealing with interruptions and distractions:
Yu J-P, Kansagra AP, Morgan J. The Radiologist's Workflow Environment: Evaluation of Disruptors and Potential Implications. JACR 2014; published online April 26, 2014
Balint BJ, Steenburg SD, Lin H, et al. Do Telephone Call Interruptions Have an Impact on Radiology Resident Diagnostic Accuracy? Academic Radiology 2014; published online September 30, 2014
In our August 2014 What’s New in the Patient Safety World column “A New Rapid Screen for Delirium in the Elderly” we discussed the importance of recognizing delirium but that delirium goes unrecognized or undiagnosed in up to 72% of cases in hospitalized patients (Collins 2010). One of the reasons may be that commonly used screening tests for delirium may not be brief enough or may require specific training for administration. We noted a new screening tool, the 4 ‘A’s’ Test (4AT) to help improve screening for delirium and its validation in a population other than that in which it was developed (Bellelli 2014).
Now another brief diagnostic tool for delirium, the 3D-CAM, has been derived and validated (Marcantonio 2014). The assessment takes only about 3 minutes to administer and can be administered by a wide variety of healthcare workers. It takes less than an hour to train someone to administer the tool.
In the validation study, the 3D-CAM had a sensitivity of 95% and specificity of 94% and performed almost equally well in patients with and without dementia (specificity in patients with dementia was slightly less at 86% but sensitivity was 96%). Importantly, the vast majority of patients identified as having delirium had either the hypoactive variety or normal psychomotor activity. That is the population in which delirium is often undiagnosed, compared to those with the hyperactive variety.
The 3D-CAM instrument and training manual are available on the Hospital Elder Life Program website. The authors note that the structured nature of the instrument also make it possible to administer via an electronic platform, such as mobile technology.
Since the study was done on general medical patients, it should be validated on other populations (eg. surgical patients, other types of hospitals, etc.) before being used in those settings.
Like the previously discussed 4AT tool, the 3D-CAM is simple, easy to administer, does not require extensive training, has good sensitivity and specificity, and works in patients with and without dementia. We fully expect that these much simpler tools should vastly improve the early detection and management of delirium.
Some of our prior columns on delirium assessment and management:
Collins N, Blanchard MR, Tookman A, Sampson EL. Detection of delirium in the acute hospital. Age Ageing 2010; 39 (1): 131-135
The 4 ‘A’s Test: screening instrument for delirium and cognitive impairment
Bellelli G, Morandi A, Davis DHJ, et al. Validation of the 4AT, a new instrument for rapid delirium screening: a study in 234 hospitalised older people. Age Ageing 2014; 43(4): 496-502
Marcantonio ER, Ngo LH, O'Connor M, et al. 3D-CAM: Derivation and Validation of a 3-Minute Diagnostic Interview for CAM-Defined Delirium: A Cross-sectional Diagnostic Test Study. Ann Intern Med 2014; 161(8): 554-561
3D-CAM (3 minute diagnostic assessment). The Hospital Elder Life Program 2014.
Go to the "Whats New Archive"
To get "What's New in the Patient Safety World"emailed to you,click here and enter "subscribe" in the subject field.
To get "What's New in the Patient Safety World"emailed to you,click here and enter "subscribe" in the subject field.