Care-seeking and antibiotic use were also analyzed separately for students who would seek a physician’s care for a common cold and those who would not Table 5. Not surprisingly, more than 50% of the students who reported that they would seek a physician’s care for a common cold endorsed seeking care in each of the presented scenarios. Likelihood of antibiotic use did not differ by symptom complex within this group, with more than 70% reporting they would use antibiotics in each of the scenarios. Among students who would not seek a physician’s help for a common cold, a significant number reported that they would still see a physician when faced with the described sets of symptoms (23% to 49% based on the scenario). High numbers of students who reported that they would not seek treatment for a common cold still reported that they were likely to use antibiotics in the presented scenarios (39% to 59% based on symptom complex).
Comparative Data
We examined the relationship of demographic characteristics to average likelihood of seeking care and average likelihood of using antibiotics (across scenarios). Reported likelihood of using antibiotics was not significantly related to any of the demographic characteristics (sex, race/ethnicity, age, college year, smoking status, or type of health insurance). Women (mean=2.89±1.14) reported being more likely than men (mean=3.17±1.19) to seek care across scenarios (P <.017). Age was also significantly related to care seeking (r=17; P <.001). The relationships between college year and care seeking and between type of health insurance and care seeking are depicted in Table 6. Freshmen and sophomores were less likely to seek care than juniors and seniors. Those using university health services were more likely than those with a private physician or other health arrangement to seek care (P <.001). Smoking status and race/ethnicity were not found to be related to reported likelihood of seeking care within this sample.
We conducted stepwise multiple regression analyses to examine predictors of belief in the effectiveness of antibiotics for cold symptoms, reported likelihood of antibiotic use across scenarios, and likelihood of seeking care across scenarios. Year of college was the only significant predictor of belief in the effectiveness of antibiotics for cold symptoms (b=0.18; P=.001). Belief in antibiotic effectiveness decreased with increasing years of higher education. Age was the only significant predictor of antibiotic use (b=0.16; P=.001). Greater age was associated with increased likelihood of using antibiotics. Average reported likelihood of seeking care was predicted by type of health insurance (b=0.22; P=.001) and by belief in the effectiveness of antibiotics to treat common colds (b=0.11; P=.03).
Discussion
A majority of students surveyed in this study would use antibiotics for the symptoms of a common cold, especially when accompanied by low-grade fever or discolored nasal discharge. This belief persisted in a significant portion even for the scenario with 5 days of clear nasal discharge and no fever. This is commensurate with other studies of different populations in the literature.13 A majority of students who reported they would not see a physician for a common cold still thought they would seek care for the presented scenarios, indicating a tendency toward inaccurate self-diagnosis especially when faced with symptoms perceived to be indicators of greater illness (eg, fever, discolored nasal discharge). Thick and opaque nasal discharge is part of the natural course of common cold and is not an indication to use antimicrobial therapy unless the symptoms persist longer than 10 to 14 days without improvement,14 but this message does not appear to be reaching patients. The use of antibiotics for viral illness and for uncomplicated URIs will lead to resistance and is discouraged by researchers and infectious disease experts.15
Our study provides evidence that even educated individuals may not recognize common URI presentations and appropriate treatment. A significant portion of this sample was unable to link their stated beliefs about colds with symptom scenarios consistent with URIs. In this sample, demographic variables did not alter the likelihood of using antibiotics, indicating the broad-based nature of the misconceptions. Our findings are consistent with previous studies showing that patients in family practice, internal medicine clinics, or emergency department waiting rooms do not recognize symptoms of the common presentation of URIs.10 The same study hypothesized that education about normal presentation could reduce visits to the physician and the misuse of antibiotics. Another recent study showed that members of ethnically diverse communities believe in the effectiveness of antibiotics for treatment of colds and often obtain them without a prescription.12
In our study, women reported a greater likelihood of care seeking across all scenarios, a finding consistent with the literature.16 Another variable predictive of seeking care was college year. Freshmen and sophomores were less likely to seek care than students in later periods of their education. This finding may have been due to the relationship between health care use and increasing age.