Management by Primary Care Clinicians of Patients Suspected of Having Community-Acquired Methicillin-Resistant Staphylococcus Aureus Infections

Univariate Analysis of SSTIs

The univariate findings for procedures and cultures were performed only on 680.x ICD-9 codes, because the majority of 681.x-682.x codes are cellulitis, for which procedures and cultures cannot be done. There was a significant decrease in the procedure rate in WHA and combined systems, but not in MCNT (Table 4). However, overall, the procedure rate was unexpectedly low, as was the culture rate. The culture rate was higher than the procedure rate, which suggests that in some cases cultures were obtained even though there was no procedure performed or the procedure was not detected in the electronic billing data.

Table 4. Pre-intervention and intervention rates for procedures and cultures of purulent SSTIs (680.x)

Key: MCNT = Medical Clinic of North Texas, WHA = Wilmington Health Associates.

In total, there were 316 cultures that were positive for S. aureus in the two systems; of these, 208 (65.8 percent) were MRSA. MRSA was highly prevalent in both systems prior to the intervention, and this did not substantially change during the intervention period. Table 5 shows the breakdown by time period and system.

Table 5. Proportion of S. aureus cultures that were MRSA

Key: MCNT = Medical Clinic of North Texas, WHA = Wilmington Health Associates.

For antibiotic usage and antibiotic choice (whether MRSA was covered when antibiotics were prescribed), the univariate analysis was performed separately on 680.x infections (Table 6 and Figure 1) and 681.x-682.x infections (Table 7 and Figure 2). This was because the antibiotics may be less indicated in general for a purulent infection that is fully drained (which may include most of 680.x cases) but, when used, should cover MRSA. For 681.x-682.x cases, of which the majority may be cellulitis infections that are not purulent, antibiotics are clearly indicated and coverage of MRSA is less imperative, according to the CDC guidelines. The combined 680.x-682.x is also presented below (Table 8). Compared to the pre-intervention period, antibiotic use increased in both systems for all ICD-9 groupings during the intervention period. The proportion of antibiotics that covered MRSA increased significantly in WHA and the two systems combined in the 681.x-682.x and all ICD-9 codes combined, whereas there was a trend for an increase in MRSA antibiotic use in MCNT within the 681.x-682.x and all ICD-9 codes combined. In the 680.x cases, there was a trend for increased MRSA antibiotic usage in WHA. Clinicians in MCNT used antibiotics more often than in WHA.

Table 6. Antibiotic usage for purulent cases (680.x) and if MRSA-covering antibiotic

Key: MCNT = Medical Clinic of North Texas, WHA = Wilmington Health Associates.

Figure 1. 680.x Skin infections and antibiotic prescribing rates in WHA, MCNT, and combined

Figure 1 is a bar chart comparing MRSA-Covering versus Not MRSA-Covering interventions and their antibiotic prescription rates. The data is shown below:

Key: MCNT = Medical Clinic of North Texas, WHA = Wilmington Health Associates

Table 7. Antibiotic usage for 681.x-682.x cases and if MRSA-covering antibiotic

Figure 2. 681.x-682.x Skin infections and antibiotic prescribing rates in WHA, MCNT, and combined

Figure 2 is a bar chart comparing MRSA-Covering versus Not MRSA-Covering interventions and their antibiotic prescription rates. The data is shown below:

Key: MCNT = Medical Clinic of North Texas, WHA = Wilmington Health Associates.

Table 8. Antibiotic usage for all SSTI (680.x-682.x) cases and if MRSA-covering antibiotic

Key: MCNT = Medical Clinic of North Texas, WHA = Wilmington Health Associates.

Return to Contents

Multivariate Analyses

The multivariate analyses adjusted for patient demographic and clinical characteristics and the findings are presented in the tables below (Tables 9-16). The intervention was associated with a trend toward a decreased likelihood of performing a drainage procedure but no significant effect on obtaining cultures (680.x cases only). The intervention increased the likelihood of prescribing antibiotics and that the antibiotics prescribed would cover MRSA.

Procedures

Table 9. Odds ratio estimates for procedures performed on purulent SSTIs (680.x)
n=436 total cases in this regression

Key: MCNT = Medical Clinic of North Texas, IM = internal medicine, FM = family medicine, PA = physician assistant, NP = nurse practitioner, Peds = pediatrics.

Cultures

Table 10. Odds ratios for cultures obtained on purulent SSTIs (680.x)
n=436 total cases in this regression

Antibiotics

Table 11. Among 680.x cases, odds ratios for receiving antibiotics
n=436 total cases in this regression

Key: MCNT = Medical Clinic of North Texas, IM = internal medicine, FM = family medicine, PA = physician assistant, NP = nurse practitioner, Peds = pediatrics.

Table 12. Among all cases (680-2.x), odds ratios for receiving antibiotics
n = 4,457 total cases in this regression

Key: MCNT = Medical Clinic of North Texas, IM = internal medicine, FM = family medicine, PA = physician assistant, NP = nurse practitioner, Peds = pediatrics, ICD-9 = International Classification of Diseases-9.

MRSA-covering antibiotics

Table 13. 680.x cases only, odds ratios for receiving a MRSA-covering antibiotic
n=436 total cases in this regression

Key: MCNT = Medical Clinic of North Texas, IM = internal medicine, FM = family medicine, PA = physician assistant, NP = nurse practitioner, Peds = pediatrics.

Table 14. Among all cases 680.x-682.x, odds ratios for receiving a MRSA-covering antibiotic
n=4,457 total cases for this regression

Key: MCNT = Medical Clinic of North Texas, IM = internal medicine, FM = family medicine, PA = physician assistant, NP = nurse practitioner, Peds = pediatrics, ICD-9 = International Classification of Diseases-9

Table 15. If an antibiotic was prescribed (680.x-682.x cases), odds ratios for receiving a MRSA-covering antibiotic (model excludes cases in which no antibiotic was prescribed)
n=1,717 total cases in this regression

Key: MCNT = Medical Clinic of North Texas, IM = internal medicine, FM = family medicine, PA = physician assistant, NP = nurse practitioner, Peds = pediatrics, ICD-9 = International Classification of Diseases-9.

Table 16. Among 681.x-682.x cases, odds ratios for prescribed antibiotics and prescribed MRSA-covering antibiotics

Key: IM = internal medicine, FM = family medicine, PA = physician assistant, NP = nurse practitioner, Peds = pediatrics.

Some other associations regarding clinicians include that pediatricians were more likely to cover MRSA, midlevels were more likely to obtain a culture, and providers at MCNT were more likely to prescribe antibiotics and more likely to cover MRSA when antibiotics were prescribed. Male patients were twice as likely to have drainage procedures done and were also twice as likely to get cultures done. Patients with diabetes were less likely to have antibiotics prescribed (680.x, cases only).

Return to Contents
Proceed to Next Section