OKPRN Projects - Completed

Mold JW, Barton ED.  OAFP starts practice-based resource/research network. J OklahomaState Med Assoc, 1996; 89: 433-434.

The Oklahoma Academy of Family Physicians in collaboration with the University of Oklahoma Health Sciences Center, has established a primary care practice-based research network with thirteen family practice offices currently participating.  The practices are connected to each other and the Health Sciences Center campuses in Oklahoma City and Tulsa by electronic mail and have begun their first research project involving the diagnosis and treatment of brown recluse spider bites in the primary care setting.

         

         

Mold JW, Cacy JR, Barton ED.  Patient-Physician E-Mail Communication.  J Oklahoma State Med Assoc, 1998 September.  91(6): 331-334.

  

Methods: 23 physicians from 14 OKPRN practices enrolled 25 consecutive patients each on two separate days (total N=1068 patients).  Each patient answered 5 questions on the back of a postal card that was then mailed to the OU DFPM. 

  

Results:

Metro

Mid-Sized

Rural

1) Do you have a computer in your home? 

55%

36%

28%

2) Do you use a computer at work?  

57%

43%

38%

3) Can you send and receive e-mail?  

48%

30%

21%

4) Would you like to use e-mail with PCP?  

68%

44%

45%

5) Do you plan to get e-mail in next 6 mo.?  

11%

7%

13%

  

 

Conclusions: A significant number of primary care patients have computer and e-mail.  Approximately half would like to be able to use e-mail to communicate with their PCP.

         

         

Cacy J, Mold JW.  The Clinical Characteristics of Brown Recluse Spider Bites Treated by Family Physicians: An OKPRN Study.  The Journal of Family Practice 1999.  48(7): 536-542.

  

Methods:  Preprinted, postage-paid cards of two types were given to physician members of the Oklahoma Physicians Research Network.  One card was used for recording examination and treatment data, and the other was for tracking the progress of the bite until it was completely healed.  For comparison, similar information was collected from the local poison control center.

  

Results:  From April 1996 to August 1998, the physicians recorded 149 bites.  Most of the victims were women aged 18 to 65 years, and most sought treatment within 3 days.  Pain was common but not universal.  Most bites were located on the extremities.  Erythema was always present, but necrosis was present in only 40% of the bites.  Some systemic signs or symptoms were common.  The most frequent forms of treatment were systemic antibiotics (66% of cases) and conservative wound management (56%).  Only 1 hospitalization occurred; 43% of the bites healed within 2 weeks; and only 13% resulted in scarring.

  

Conclusions:  The brown recluse spider bite is a relatively common problem treated by family physicians in Oklahoma .   Most bites are not serious.  More research at the practice level is needed to develop diagnostic criteria and to explore the effectiveness of various treatment options.

         

         

Mold JW. Treatment of brown recluse spider bites by family physicians.  Submitted to JABPF (in press)

         

Methods: Same as for first brown recluse study.  Total N 256 bites.

  

Results: Predictors of longer healing time:Delay in seeking treatment. Diameter of erythema when first seen. Diameter of necrosis when first seen. Diameter of necrosis when first seenTreatment with dapsone, systemic steroids, or high dose Vitamin C. Predictors of scarring: Diameter of necrosis when first seen. Diabetes mellitusDiameter of necrosis when first seenDiabetes mellitusTreatment with dapsone or high dose Vitamin C

  

Conclusions: None of the treatment approaches used by OKPRN physicians made any difference in healing time or probability of scarring.  Fortunately most bites (80%) healed without scarring.  The median time to healing was 17 days.

         

         

Mold JW and Thompson DM.  Management of brown recluse spider bites in primary care.  JABPF September/October 2004, 17(5): 347-352.

         

         

Brown Recluse Spiders Unpublished

         

Background and Rationale: Little is known about the level of knowledge and experience of Oklahomans with regard to brown recluse spider bites.  There may be a need for additional public education on this subject. 

         

Methods: All adult (>18 y.o.) patients being seen in the waiting rooms of the Family Medicine Center clinics and individuals shopping at a grocery store in Edmond during predetermined periods of time were approached and asked to answer a series of questions and to pick out a picture of a brown recluse spider from among five pictures of spiders. 

         

Results: 132 FMC patients and 90 shoppers completed the survey.  Although the mean ages (43.6 vs. 21.4) and gender distribution (67% female vs. 50% female) of the two groups were different, their responses to the survey were practically identical.  46% thought they could identify a BR spider and 44% chose the correct picture.  Twenty-four percent had seen a BR spider in their home in the past year, but only 2% had been bitten by one.   

         

Conclusion: Brown recluse spiders are common in Oklahoma , but less than 50% of a sample of Oklahomans can recognize them from a picture.  Considering their prevalence, BR spiders only rarely seem to bite people.

         

         

Hulson TD, Mold JW, Scheid D, Aaron M, Aspy CB, Ballard NL, Boren N, Gregory ME, Truong TC.  Diagnosing Influenza: The Value of Clinical Clues and Laboratory Tests.  The Journal of Family Practice 2001 December.  50(12): 1051-1056. 

         

Methods:  Data were collected during 3 consecutive influenza outbreaks over a 2-year period in one 5-man family practice office in Edmond .  The information collected included date of onset, symptoms, vaccine status, WBC and differential counts, ZstatFlu test (antigen test), and influenza culture.  Using culture positivity as the criterion for influenza diagnosis, we compared cases with non-cases on each variable independently and by logistic regression. 

         

Results:  Culture-positive cases could not be reliably distinguished from those that were culture negative using symptoms or vaccination status.  Both WBC count and ZstatFlu results discriminated fairly well, and their combination did somewhat better.  Differential counts were not helpful.  WBC counts above 8,000 were associated with a low probability of influenza.  The sensitivity and specificity of the ZstatFlu were 65% and 83%, respectively. 

         

Conclusions:  Our data suggest that symptoms and vaccine status do not reliably identify patients with influenza.  Use of WBC counts and the ZstatFlu test can be helpful.  The sequence, combination, and criteria for use of these tests depend on tradeoffs between under-treatment of influenza cases and the over-treatment of non-influenza cases, and cost and benefit projections for individual patients.

         

         

Mold JW, Gregory ME.  Best Practices Research.  Family Medicine 2003 February, 35(2): 131-134. 

         

This is a description of the “best practices research” method pioneered for use in practice-based research networks by OKPRN.  It is a method used to identify effective methods for managing common primary care processes like lab test results, prescription refills, no-shows, preventive services, etc.  The process of interest is first broken down into to its component parts.  Values and acceptable standards are then agreed upon for each part (to define “most effective”).  Finally, exemplars are sought and identified from within the network for each part, their methods are studied, and the best methods are combined into a complete “best” method.

         

         

Mold JW, Cacy DS, Dalbir DK.  Management of Laboratory Test Results in Family Practice.  An OKPRN Study.  The Journal of Family Practice, 2000 August.  49(8): 709-715. 

         

Methods:  We used a questionnaire guided by a literature review to identify a conceptual model, current practices, and clinicians who reported having an effective method for at least one of 4 steps in the process of managing laboratory test results.  Clinicians with differing methods were selected for each of the steps.  Practice audits and patient surveys were used to determine actual performance.  On the basis of these audits, we constructed a unified best method and conducted time-motion studies to determine its cost. 

         

Results:  After auditing only 4 practices we were able to identify effective methods for 3 of the 4 steps involved in the management of laboratory test results.  The unified best method costs approximately $5.19 per set of tests for an individual patient. 

         

Conclusions:  By identifying effective practices within a family practice research network, an effective method was identified for 3 or the 4 steps involved in the management of laboratory test results in primary care settings.

         

         

Mold JW, Enright MM, Oehlert WH.  Improving the quality of care for diabetic patients using best practices research and practice enhancement assistants.  In preparation. 

         

Methods: Exemplars identified from previous OFMQ audits were interviewed regarding their diabetes management methods.  Six common principles were discovered.  These principles were then taught to 20 OKPRN physicians who were also provided with a practice enhancement assistant to help them make changes.  A PDA-based program was also available to use by choice.  Pre- and 3 mo. Post-intervention audits were done.

  

Results:

Pre

Post

p-value

Annual A1c 

87%

96%

0.0003

Annual lipid profile

69%

80%

0.02

Annual urine protein

53%

64%

0.05

Annual foot exam

71%

82%

0.004

Annual eye exam ref

48%

59%

0.04

Annual eye exam doc

33%

35%

NS

ACE-I for HTN

72%

86%

0.03

ACE-I for proteinuria

74%

94%

0.007

Pneumovax ever

42%

61%

0.0006

Annual flu shot

58%

67%

0.03

Mammography Q2y

69%

68%

NS

  

Conclusions: Use of the best practices research approach in combination with a PEA resulted in significant short-term improvements in quality of diabetes care.

         

         

Roberts M and Mold JW. Primary care network survey.  Presented at OKPRN Convocation 2/2003. 

         

Methods: The PRINS surveys were designed by the Agency for Healthcare Research and Quality and primary care practice-based research network directors.  They were intended to capture information on the clinicians and practices involved in primary care practice-based research networks and a sample of their patient encounters, and to compare them to other clinicians and practices using other available data sets.  Overall, 1500 PRINS 1 surveys were completed by clinicians in 19 networks across the country, and PRINS 2 information was captured on 26,649 encounters.  In OKPRN, a total of 248 clinicians, including 98 faculty members, 100 second and third year residents, and 50 community practitioners, completed the PRINS 1 survey (clinician and practice information), and 40 faculty, 52 residents, and 26 community practitioners completed data on approximately 30 encounters each, capturing data on a total of 3535 encounters.  

         

Results: PRINS 1: (OKPRN compared to Family Physicians nationally)Community Practitioners: OKPRN practitioners were more likely to be younger, male, and Native American; they were less likely to be Hispanic and black.  Faculty: OKPRN faculty were more likely to be younger and female.Residents: OKPRN resident members were younger, female, and Native American; they were less likely to be African American.PRINS 2: (OKPRN compared to PCPs in the National Ambulatory Medical Care Survey)Community Practitioners: On average, OKPRN physicians see more Native Americans and Medicare patients, and fewer Hispanics, Medicaid, and HMO patients.  They do less non-illness care and more ambulatory surgery.   Faculty: OKPRN faculty physicians see a higher percentage of female, Native American, Medicaid, and HMO patients and fewer Hispanic patients.  They do less non-illness care and more ambulatory surgery than PCPs surveyed for NAMCS.Residents: OKPRN residents see more female, Black, Native American, Medicaid, self pay, and HMO patients and many fewer Caucasian and Medicare patients.  They do more non-illness care than NAMCS PCPs. 

         

Conclusions: While OKPRN clinicians are similar to national samples, they differ in several important ways.  Within OKPRN, residents see a very different spectrum of patients than faculty and particularly community physicians do.

         

         

Mold JW, Mathew MK, Belgore S, DeHaven M.  Prevalence of night sweats in primary care patients.  An OKPRN and TAFP-Net collaborative study.  The Journal of Family Practice 2002 May, 51(5): 452- 456. 

         

Methods:  OKLAHOMA Studies pilot year data.  799 patients 65 years of age and older from the practices of 23 OKPRN physicians within 75 miles of OKC.  Patients completed and extensive questionnaire that was reviewed with them by a research nurse. 

         

Results:  Of the 2267 patients who participated, 41% reported experiencing night sweats within the last month, including 23% with pure night sweats and an additional 18% with day and night sweats.  The prevalence of night sweats in both men and women was highest in the group aged 41 years to 55 years.  In multivariate analyses, factors associated with pure night sweats in women were hot flashes and panic attacks; in men, sleep problems.  Variables associated with night and day sweats in women were increased weight, hot flashes, sleep disturbances and use of antihistamines, selective serotonin reuptake inhibitors (SSRIs) and other (non-SSRI, non-tricyclic) antidepressants; in men, increased weight, hot flashes, and greater alcohol use.   A majority of patients had not reported their night sweats to their physicians, even when frequent and severe. 

         

Conclusions:  Night sweats are common and under-reported.  Pure night sweats and night and day sweats may have different causes.  With regard to the etiologies of pure night sweats, panic attacks and sleep disorders need further investigation.

         

         

Mold JW, Aboshady HM, and Roberts AM. Prevalence and predictors of night sweats, day sweats, and hot flashes.  Annals of Family Medicine 2004 September/October, 2(5): 391-397 

         

Methods:  A cross-sectional study of patients over 64 years old recruited from the practices of 23 family physicians.  Variables included socio-demographic information, health habits, chronic medical problems, symptoms, quality of life (QOL), and the degree to which patients were bothered by night sweats, daytime sweating, and hot flashes. 

         

Results:  Among the 795 patients, 10% reported being bothered by night sweats, 9% day sweats, and 8% hot flashes.  Eighteen percent reported at least