Developing a Career in Patient Safety and Quality Improvement in Ambulatory Care (Text Version)

On September 14, 2009, Amanda G. Kennedy made this presentation at the 2009 Annual Conference. Select to access the PowerPoint® presentation (247 KB) (Plugin Software Help).

Slide 1

Developing a Career in Patient Safety & Quality Improvement in Ambulatory Care

Amanda G. Kennedy, PharmD, BCPS
University of Vermont
AHRQ K08 HS013891

Slide 2

Broad Career Goals

• Develop skills in quantitative and qualitative approaches
• Understand the challenges of conducting research in ambulatory environments
• Pose and answer patient safety and quality improvement research questions that are FINER*
• Share findings broadly
• Obtain funding to continue research in primary care

*From: Hulley SB. Designing Clinical Research 3rd Ed.

Slide 3

Goals for Today

• Present two case examples of patient safety research projects funded by AHRQ K08 that incorporate broad career goals
• Share lessons learned / policy implications

Slide 4

Case #1: Research Question

• Can a prescribing error reporting system be implemented and sustained in primary care using existing office systems?
  • Is it feasible?
  • Will people report?
  • What kind of data can be obtained from existing office tools and will the data be useful?
  • Is it sustainable?

Kennedy AG, Littenberg B, Senders JW. Using nurses and office staff to report prescribing errors in primary care. Int J Qual Health Care 2008 Aug; 20(4):238-45. PMID: 18430748

Slide 5

Methods

• 7 primary care practices in Vermont
• 103 prescribers, managers, nurses and admin staff
• Practice encouraged to report communications with community pharmacists regarding prescription problems for 6 months
• Practice was encouraged to use their usual methods for documenting pharmacist communications as the reports

Slide 6

Analysis

• All reports were classified by:
  • Severity (based on NCC MERP)
  • Setting (practice, pharmacy, patient)
  • Mode (omission, commission, no error, indeterminate)
  • Prescription domain (drug, strength, route, etc.)
  • Error-producing conditions (environmental, team, individual or task factors that affect performance)*
• Moderate agreement in analyzing data (kappa 0.552, standard error 0.044, P, 0.001)

Reason J. Human Error 1990. Dean B, et al. Lancet 2002;359:1373�8.

Slide 7

Results

• Is a primary care prescribing error reporting system feasible? - YES
  • All practices submitted reports
  • Total reports per practice varied from 10 to 62 reports (median 32 reports per practice)
• Will people report? - YES
  • Nurses and office staff contributed most
  • Prescribers contributed only 7 reports (3.5%)

Slide 8

Results

• What kind of data can be obtained from existing office tools and will the data be useful?
  • Antidepressants (38/216), narcotics (32/216) and antihypertensives (24/216) most frequently reported
  • Bupropion was the individual drug most often reported (12/216), followed by levothyroxine (6/216) and metoprolol (6/216)
  • 20% of near-misses or errors (43/216) concerned 'high-alert' meds
• Is it sustainable?  NO
  • Reporting decreased by 3.6 reports per month
    • (95% CI, 22.7 to 24.4, P<0.001)
• No continued reporting

Slide 9

Lessons Learned: Reporting System

• Existing primary care systems for handling communications with pharmacists are highly variable
• Primary care practices are willing to report prescribing errors but are distracted by competing priorities
• Building on existing systems for reporting within a practice is feasible and causes minimal disruptions, but is not sustainable
• Existing systems can capture simple descriptions of prescribing errors and some basic insights into mechanism of error, but do not contain rich descriptions
• Integrating/automating reporting may increase sustainability as long as there is feedback to staff AND providers

Slide 10

Lessons Learned: Prescribing Errors

• Can be fixed with e-prescribing
  • Prescribing for strengths not commercially available
  • Illegibility
  • Omission errors
• Probably cannot be fixed with e-prescribing
  • Multiple formulation issues (XR, XT, XL, ER, etc)
  • Patient use of multiple providers and pharmacies
  • Med selection issues (look-alike, sound-alike)

Slide 11

Case # 2: Research Question

• What are the perceived causes (or contributors) of prescribing errors by primary care providers, pharmacists, and pharmacy technicians?

Slide 12

Methods

• Prepared and piloted interview guide
• Facilitated 10 focus groups
  • 3 physician
  • 1 nurse practitioner
  • 3 pharmacist
  • 3 pharmacy technician

Slide 13

Analysis

• Transcript-based
• Used a multidisciplinary panel
• Framework: Reason's Accident Causation Model
• Coded in Nvivo 7

Slide 14

Results: Upper-level Decisions

• The Joint Commission: "There is a JCAHO requirement that says they have to document that they have checked the medications. And, it is interesting to me because this has actually come up and we keep saying, that's great that they have to do it. But, it's not right half of the time. Nobody seems to actually care about that." Prescriber
• Corporate Management: "We're working on one pharmacist and they make promises like when we reach 250 [prescriptions per day] we get another pharmacist and we reached that for quite awhile and they said no now it will be 300. Now we have way surpassed that and we are still working with one pharmacist." Pharmacy Technician

Slide 15

Insurance Companies

• "Many times you are challenged because the patient's on a specific plan or the patient is only eligible for certain things. And, so you are taking the same prescription and being asked to do something different and you've already put your thought process into the first prescription and now it's either an even trade and you don't really care or maybe it isn't really an even trade but you're not thinking about that any more because it is out of context with the visit completely and you're just sort of being challenged to do something else. So, I think that can lead to an error." Prescriber
• "They put the prescription into their system and it kicks back out and says you can't prescribe that drug or it's a $30 co-pay as opposed to the $2 co-pay. Then they have to tell the patient your prescription plan does not pay for that drug, the drug your doctor prescribed. Now then it bumps back to our office. At that level, then the documentation is if you call in a different drug then it gets documented again. So, there is a paper trail for it but the amount of time spent with multiple interactions between different entities: the drug management company, the pharmacist, the physician, the patient. None of that time or confusion is documented any where other than in our psyches." - Prescriber
• "What's happening is perhaps 30% of each pharmacist's day is spent resolving insurance issues. And what used to be an individual that is accessible to the public aren't as readily accessible any more because they're on the phone with insurance companies." - Pharmacist

Slide 16

Lack of standard protocols & ownership

• "As your patients get more complex they have more hands in the brew. And, additions and deletions and modifications of prescriptions occur outside of your own office and patients are not always adequate to give you the testimony about that. They don't bring their bag of medicines to every visit so then you're not always sure that they are on the dose of this that you think they're on. But, you find out about it a month later when they call your office for the refill and it's like they're on what?" � Prescriber
• "Depending on the medication and depending on if it's for somebody else's patient and what the medication is, depends on how much I'm going to dig. And the nurse, depending on who she is, might have brought me the chart or might not have. If she hasn't then I need to go look for the chart and maybe I'll go do that or maybe I'll be busy and maybe I'll take a little more leeway than I might normally if the chart was given to me." Prescriber

Slide 17

Lack of standard protocols & ownership

• "Phone-ins are an issue. A lot of times they don't know how to say a word and then you're thinking how do I know this is right? And especially if it is something that's kind of odd. And so you know, the difficulty there is you question that person and they either say well that's what it says or that's what the doctor wrote or I'll have to get back to you." - Pharmacist
• "I feel like you rarely ever get to talk to any physicians and even when you really need to talk to the doctor it's impossible. The front office staff thinks that you wanting to talk to the doctor is kind of a joke." Pharmacist

Slide 18

Lessons Learned

• Many perceived causes or factors influencing prescribing errors are latent*
  • Upper-level decisions (e.g. Formularies, policies)
  • Lack of standards and protocols for managing a prescription throughout the prescription process, especially when there is a problem (i.e. Who "owns" the prescription?)
  • It is unlikely that individual providers or healthcare workers can solve these problems without policy changes, standards, and accountability

*Reason J. Human Error 1990. Dean B, et al. Lancet 2002;359:13738.

Slide 19

Summary and Conclusions

• Challenges for studying & implementing prescription safety improvements in ambulatory care
  • Silos
  • Competing priorities
  • Lack of standard protocols across practices
• Cautious Optimism
  • Provider interest in quality and safety
  • E-prescribing
  • New delivery models to reduce silos (e.g. patient centered medical home)

Slide 20

Thank you

• K08 Mentors
  • Benjamin Littenberg, MD (Vermont)
  • John W. Senders, PhD (Toronto)
• Focus Group Analysis Team
  • Benjamin Littenberg, MD
  • Laurie Hurowitz, PhD
  • Kairn Kelley, PhD Candidate
  • Rodger Kessler, PhD
  • Jennifer Otten, PhD
  • Richard Pinckney, MD, MPH
  • Jennifer Prue, EdD
  • Alan Rubin, MD