Relation between general practitioners' prescribing of antibacterial drugs and their use of laboratory tests

BMJ 1996;313:922 (12 October)

General practice

Relation between general practitioners' prescribing of antibacterial drugs and their use of laboratory tests

Michael C Kelsey, consultant microbiologist,^a George A Kouloumas, medical adviser in prescribing,^b Peter A Lamport, senior chief biomedical scientist,^a Cheryl L Davis, pharmaceutical adviser ^b

^a Department of Microbiology, Whittington Hospital, London N19 5NF, ^b Camden and Islington Family Health Services Authority, 5th Floor, Insull Wing, London NW1 2LJ

Correspondence to: Dr Kelsey.

Only doctors can prescribe antibacterial drugs, and they musttake responsibility for prescribing effective, appropriate,safe, and economic drugs. We audited the prescribing habitsof general practices and related these to a list of antibioticsselected as first line drugs by a group of general practitioners.Compliance with this list was assumed to represent good practice.General practitioners have unrestricted access to facilitiesfor laboratory testing of suspected infections, and we studiedthe association between practices achieving a high standardof prescribing antibiotics and their appropriate requests forinvestigations of children's urinary tract infections and women'sgenital infections.

Methods

and results

Forty one general practitioners attended prescribing forumsand formulated a list of first line antibacterial drugs (seefootnote of table 1) which they thought would meet all theirprescribing requirements not guided by microbiological sensitivitydata. In our study we regarded good compliance as 85% of a practice'santibacterial prescriptions being met from this list.

Table 1--Comparison of general practices' demographics and use of microbiology
laboratory by compliance with list of recommended antibiotics* (values are medians
(ranges) (95% confidence intervals) unless stated otherwise)
----------------------------------------------------------------------------------------------------------------
                                                                     Compliance with list
----------------------------------------------------------------------------------------------------------------
                                                              Poor (<85%)         Good (>/=85%)       P value+
----------------------------------------------------------------------------------------------------------------
No of patients registered in practices                       4245 (1689-8277)   4611 (2007-11572)
                                                              (3044 to 5490)     (4162 to 7219)         0.18
% Of female patients in practice                                  49.26              49.53
                                                              (43.93-56.13)      (43.73-54.70)
                                                             (47.35 to 51.53)   (47.96 to 50.91)        1.00
% Of patients aged <5 years in practice                      4.81 (2.90-8.22)    5.68 (3.53-8.98)
                                                              (3.99 to 5.74)     (5.26 to 6.78)         0.05
No of ASTRO-PUs                                                   13947              15172
                                                              (4655-26281)       (5957-35550)
                                                             (10124 to 16916)   (13192 to 22684)        0.32
No of antibacterial prescriptions per ASTRO-PU               0.22 (0.13-0.32)   0.22 (0.16-0.32)
                                                              (0.19 to 0.26)     (0.19 to 0.23)         0.75
Cost of antibiotics per ASTRO-PU (£)         0.75 (0.28-2.01)   0.62 (0.36-1.40)
                                                              (0.63 to 1.23)     (0.54 to 0.81)         0.22
Cost per prescription for antibiotics (£)    3.13 (2.11-8.87)   2.91 (1.82-5.44)
                                                              (2.88 to 5.45)     (2.66 to 3.66)         0.21
Urine samples sent for investigation:
  No of samples per 100 patients                             1.5 (0.02-9.65)    3.5 (0.5-10.95)
                                                              (0.82 to 4.06)     (3.14 to 6.03)         0.01
  No of samples per 100 patients aged <5                       1.0 (0-4.53)      4.5 (0.4-11.2)
                                                             (0.64 to 2.71)      (3.14 to 7.10)         0.003
  % Of antibiotic tests showing resistance++                 14.6 (11.2-29.7)   14.2 (8.5-30.0)
                                                              (11.5 to 21.6)     (12.6 to 17.3)         0.95
Genital swabs from women aged 15-45 sent
    for investigation:
  No of high vaginal swabs per 100 women                      0.4 (0-15.9)       4.0 (0-13.9)
                                                             (0.40 to 4.49)     (3.14 to 7.10)          0.004
  Ratio of cervical swabs to high vaginal swabs                0 (0-0.26)       0.01 (0-0.73)
                                                             (0.02 to 0.08)     (0.03 to 0.26)          0.04
----------------------------------------------------------------------------------------------------------------
* Antibiotics recommended as first line treatments: penicillin V, amoxycillin, flucloxacillin, co-amoxiclav,
erythromycin, trimethoprim, oxytetracycline, doxycycline, metronidazole, and cephalexin.
+Mann-Whitney U test.
++Practices with more than 10 isolates.

We studied 32 general practices that principally used WhittingtonHospital's microbiology laboratory. We retrieved details ofall requests submitted by these practices from the laboratory'sdata systems. We selected all the prescribing data for antibacterialdrugs (as defined in British National Formulary¹) for April1993 to March 1994 from prescribing analysis and cost (PACT)records. We obtained data on list sizes and demographic detailsfrom the health authority's records, and we calculated a standardisednumber of prescribing units (ASTRO-PUs) for each practice usingthe method of Roberts and Harris.²

The 32 practices covered 160 981 patients. Eighteen of the practicesachieved at least 85% compliance with the list of selected antibiotics.The median level of compliance was 87% (95% confidence interval80% to 87%). The number of prescriptions issued, the cost perASTRO-PU, and the prescription cost did not differ between thepractices that achieved good compliance and those that did not(table 1). The poorly complying group treated fewer childrenaged under 5 years (P = 0.05). Of the 16684 requests for bacteriologicalexamination, 8130 were for urine samples (from 6460 patients),and 700 of these were from children aged under 5. Practiceswith good compliance submitted significantly more urine samplesper 100 patients as well as per 100 children aged under 5 (table1). Five practices submitted no urine samples from patientsaged under 5, and all achieved less than 85% compliance. A totalof 16 703 sensitivity tests were performed on 1720 isolates.There was no difference in the proportion of isolates with antibioticresistance between the two groups of practices (table 1).

Practices with good compliance submitted significantly morehigh vaginal swabs, and the significantly higher ratio of cervicalswabs to high vaginal swabs in these practices indicated thatthey were more likely to send samples from both vagina and endocervix(table 1). The level of performance, as indicated by the submissionof urine samples and high vaginal swabs from each practice,was highly correlated (Spearman's coefficient of correlation0.8558, P<0.001).

Comment

The general practitioners' compliance with prescribing guidelinescorresponded with other standards of quality. Confirmation ofurinary tract infection in a feverish child requires microbiologicalinvestigation of a urine sample.³ Significantly fewer children'surine samples were submitted by the less compliant practices,and five practices did not submit any samples from childrenunder 5. The use of urine dip sticks with nitrite and leucocyteesterase tests is not an adequate replacement for culture: therate of false negative results for nitrite can be up to 48%.⁴Microscopy for pyuria also has a substantial false negativerate.⁵

Both vaginal and endocervical infections must be diagnosed insexually active women. The practices that underinvestigatedurinary tract infections in children also submitted fewer genitalsamples from women, and this poor performance with genital sampleswas also related to antibacterial prescribing. Neither the costnor the number of prescriptions issued related significantlyto compliance with the prescribing guidelines.

Funding: None.

Conflict of interest: None.