Reducing inappropriate antibiotics prescribing: The role of online commentary on physical examination findings

Abstract

Objective

This study investigates the relationship of ‘online commentary’ (contemporaneous physician comments about physical examination [PE] findings) with (i) parent questioning of the treatment recommendation and (ii) inappropriate antibiotic prescribing.

Methods

A nested cross-sectional study of 522 encounters motivated by upper respiratory symptoms in 27 California pediatric practices (38 pediatricians). Physicians completed a post-visit survey regarding physical examination findings, diagnosis, treatment, and whether they perceived the parent as expecting an antibiotic. Taped encounters were coded for ‘problem’ online commentary (PE findings discussed as significant or clearly abnormal) and ‘no problem’ online commentary (PE findings discussed reassuringly as normal or insignificant).

Results

Online commentary during the PE occurred in 71% of visits with viral diagnoses (n = 261). Compared to similar cases with ‘no problem’ online commentary, ‘problem’ comments were associated with a 13% greater probability of parents questioning a non-antibiotic treatment plan (95% CI 0–26%, p = .05,) and a 27% (95% CI: 2–52%, p < .05) greater probability of an inappropriate antibiotic prescription.

Conclusion

With viral illnesses, problematic online comments are associated with more pediatrician–parent conflict over non-antibiotic treatment recommendations. This may increase inappropriate antibiotic prescribing.

Practice implications

In viral cases, physicians should consider avoiding the use of problematic online commentary.

Introduction

Inappropriate antibiotic prescribing for viral upper respiratory tract infections (URI) is common in the United States [1]. Nearly one-third of patients diagnosed with the common cold receive an antibiotic prescription, and prescribing for bronchitis and other viral illnesses are estimated to be higher than 50% [2], [3]. While a number of recent studies indicate that inappropriate antibiotic prescribing may have peaked in the 1990s [4], [5], [6], [7], [8], [9], [10], [11], there is evidence to suggest that declining prescribing rates are primarily driven by decreased rates of office visits for respiratory tract infections [8], [12]. For patients presenting with URIs, physicians’ rates of prescribing show a relatively modest and uneven decrease [8], [12], together with increased reliance on broad-spectrum agents [1], [4], [10]. Inappropriate use of antibiotics has led to increased resistance among many strains of bacteria that commonly infect both children and adults [13], [14], [15], [16], [17], posing risks both to the individual and the community [13], [15], [18], [19], [20].

Upper respiratory tract infections are a common reason that parents seek medical care for illness in their children [3], [21], and account for approximately 75% of children's antibiotic prescriptions [8]. Although 50–70% of parents report a pre-visit expectation that their child will be given antibiotics [22], [23], 65–70% of these infections are viral and thus cannot effectively be so treated [24], [25], [26], [27], [28], [29].

We previously identified a communication practice in which physicians convey their physical examination findings during, as opposed to after, the physical exam [30]. We call this communication practice online commentary. We identified two main types of online commentary: no problem online and problem online (see Table 1). In no problem online, the physician provides reassurance by discussing normal findings on the physical examination (PE) or by suggesting that positive findings are minor and/or not significant. In problem online, PE findings are reported as being clearly abnormal and/or are evaluated as significant. We hypothesized that ‘no problem’ online commentary would be associated with a reduction in the probability of inappropriate prescribing [30]. Correspondingly, we also hypothesized that ‘problem’ online commentary would be associated with an increase in the probability of inappropriate prescribing, and that offering no commentary on physical examination findings would be neutral in its effects.

In a subsequent paper, we found some evidence that when children had viral illnesses, physicians who used problematic online commentary inappropriately prescribed antibiotics more often than physicians who exclusively used no problem online [31]. However, because this prior work was done with a relatively small, homogenous population of parents and physicians, the results were suggestive rather than conclusive.

Our previous work also demonstrated that when physicians perceived parents as expecting antibiotics they were significantly more likely to inappropriately prescribe them for viral diagnoses [23]. In a prior analysis of data from the current study, we found that physicians had a higher probability of perceiving a parent as expecting antibiotics if the parent questioned the physician's non-antibiotic treatment plan for her child [32]. We also found that a key determinant of parents questioning the physician's treatment plan was the physician ruling out the need for antibiotics when discussing the plan [32]. These prior findings led us to question whether using problematic online commentary in the context of an ultimately viral diagnosis might also increase the probability that a parent would question the physician's treatment plan.

We theorized that upon a viral diagnosis, a parent would not question a non-antibiotic treatment plan if the physician had made only reassuring comments about the child's physical exam (‘no problem’ online) or made no comments about the exam at all (no online). In contrast, we theorized that describing a child's physical examination findings as concerning (‘problem’ online), but then assigning a diagnosis that did not warrant antibiotic treatment, might result in cognitive dissonance for the parent [33], and thus increase the probability that she would question a non-antibiotic treatment plan.

The primary goals of the current study are to examine the relationships among physician online commentary use, parent questioning of the physician's treatment plan, and the physician's ultimate prescribing decision. We hypothesize that in viral cases, if the physician used problematic online commentary, the parent would more often question the physician's non-antibiotic treatment plan. We further hypothesize that problematic online commentary in viral cases would be associated with higher rates of inappropriate prescribing above and beyond the increase associated with physician perceptions that parents expect antibiotics.