Counseling Smoking Parents of Young Children: Comparison of Pediatricians and Family Physicians

Background  Secondhand smoke is a major cause of morbidity in young children, and exposure to smoking parents is the principal source. Physician visits for young children present an opportunity to effect behavioral change among smoking parents.

Objective  To survey pediatricians and family physicians in their knowledge and practice of smoking cessation counseling with parents.

Design  Cross-sectional mail survey.

Setting  Urban California.

Participants  Pediatricians and family physicians in urban areas of California, younger than 65 years, practicing in an ambulatory setting, and randomly selected from the American Medical Association Physician Masterfile.

Main Outcome Measures  Reported frequency of asking about tobacco use, using cessation counseling techniques with smokers, and perceived barriers to providing cessation services.

Results  Of the 1000 mailed surveys, 899 were eligible and 499 (56% response rate) were returned and completed. A higher proportion of pediatricians compared with family physicians were women (44% vs 29%; P<.01) and nonwhite (44% vs 32%; P = .01). Family physicians compared with pediatricians were more likely to report referring a parent to a smoking cessation program (41% vs 30%), giving pamphlets on smoking cessation (40% vs 28%), asking for a quit date (41% vs 18%), scheduling a follow-up visit to discuss quitting (27% vs 5%), and recommending nicotine replacement therapy (41% vs 13%) (for each comparison, P<.001). Pediatricians were more likely to report recording in the medical record smoking by a parent as a problem for the child (65% vs 48%; P<.001), but a higher proportion of pediatricians perceived that parents would ignore the advice (39% vs 24%; P<.001) and lacked interest in quitting smoking (45% vs 27%; P<.001). Pediatricians were more likely to agree that they lacked smoking cessation counseling skills (26% vs 7%; P<.001). Multivariate models showed that pediatricians were less likely to report performing 5 of 14 smoking cessation techniques in at least 50% of smoking parents.

Conclusions  Pediatricians appear to lack training to implement smoking cessation counseling with smoking parents. Physicians in private practice are less likely to counsel smoking parents. Educational interventions for pediatricians are needed to decrease secondhand smoke exposure for young children.

SECONDHAND SMOKE exposure is associated with a significant increase in morbidity and mortality among children, especially those younger than 5 years.^1-3 In 1986, an estimated 53% to 76% of households contained one adult resident who smokes and up to 12 million children aged 5 years and younger were exposed to environmental tobacco smoke.⁴ In California, 35% of children aged 5 years and younger were exposed to environmental tobacco smoke for an average of about 4 hours per day.³

Based on the data supporting implementation of clinician-mediated interventions to promote smoking cessation among adults,⁵ the National Cancer Institute developed a program that targets clinicians caring for children.⁶ The strategy to incorporate primary care clinicians in a public health approach to smoking cessation has been strengthened by the development of evidence-based guidelines.⁷ However, in contrast to the data on physicians caring for adults, there are limited studies^8-12 that have evaluated the effectiveness of this program or the usefulness of interventions focused on decreasing secondhand smoke exposure for children.

Several studies have evaluated whether pediatric clinicians are able to assist parents to quit smoking and help decrease secondhand smoke exposure for young children. In a survey¹³ of residents in the 3 primary care specialties, only 32% of pediatric residents reported any training in smoking cessation counseling and most scored significantly lower compared with family medicine and primary care internal medicine residents. Two surveys^14,15 of practicing pediatricians in Maine and Vermont showed that 91% to 94% advise smoking parents to quit, spending an average of 4 to 5 minutes doing so, but only about 10% recorded the information in the medical record. However, less than half of these pediatricians talked to most parents about the effects of smoking on their children.^14,15 Secondhand smoke was addressed by the Vermont pediatricians by advising parents to smoke away from children (77%), to reduce consumption of cigarette use (40%), and to set a quit date (22%).¹⁵

Parents of young children visit pediatricians more often than any other physician, and these visits are an opportunity to effect behavioral change among smoking parents.^14,15 Pediatric clinicians may be uniquely positioned to counsel these adults to quit smoking, but they must be comfortable with the topic. To evaluate and compare the knowledge, attitudes, and practices with respect to counseling parents of children younger than 5 years about secondhand smoke, we conducted a survey of California pediatricians and family physicians. We selected family physicians as a comparison group who also take care of children,¹⁶ but would be expected to use the smoking cessation techniques from adult practice. We specifically asked about the physicians' practice in delivering smoking cessation interventions to these parents and the perceived barriers in counseling parents to quit smoking. Our ultimate goal is to develop and implement an educational intervention directed at pediatricians and family physicians to promote smoking cessation among smoking parents of young children.

From the 1997 American Medical Association Physician Masterfile data set for California, we selected a stratified random sample of 1000 pediatricians or family physicians, aged 65 years or younger, practicing in urban areas of California. After the surveys were returned, we retained data only for physicians who continued to be clinically active, saw patients aged 18 years and younger, and spent time practicing in ambulatory care settings.

We developed the survey to ascertain how often physicians assess smoking behavior of parents, their practices regarding counseling smoking parents to quit, and perceived barriers to the delivery of smoking cessation counseling to these parents. Physicians were asked to estimate the percentage of parents or accompanying adults who were asked, during a routine visit, if they smoke tobacco regularly; if they or anyone else smoke in the home; and whether a smoking parent smokes in front of their child outside the home. Physicians were asked to estimate the frequency (0%-100% of the time) they provided 14 possible smoking cessation practices during routine visits.

The 14 practices were as follows: discussing a child's health risk from secondhand smoke, advising a smoking parent to quit, counseling on the health risks of smoking, recording smoking in the medical record as a health problem for the child, telling parents that smoking is a bad example for the child, increasing the motivation of a parent to quit smoking, referring parents to their personal physician for smoking cessation counseling, referring parents to a smoking cessation program, providing pamphlets on self-help smoking cessation, asking parents if they are willing to set a quit date, prescribing or recommending use of nicotine replacement therapy to quit, asking a nurse in the office to counsel parents to quit smoking, scheduling a follow-up visit or telephone call to discuss quitting, and telephoning parents to discuss progress with quitting. Four of these 14 practices have been central components in randomized trial studies that have demonstrated increased abstinence among smoking patients when clinicians used these practices. These effective practices included setting a quit date, prescribing nicotine replacement therapy, scheduling a follow-up visit or telephone call, and providing nurse-mediated counseling in the office.⁷

Respondents were also asked to estimate the amount of time spent counseling smoking parents on tobacco cessation in categories of minutes (0, 1-2, 3-5, 6-10, and ≥11 minutes). Finally, we asked the respondents about their level of agreement with 9 perceived barriers to the delivery of these smoking cessation services (eg, If I ask all parents whether they smoke tobacco, it would anger parents). The questionnaire also contained items about the respondents' demographics, history of smoking cigarettes, medical education history, and type of practice. The questionnaire was developed by the investigators (E.J.P.-S., M.J.-R., C.P.K., E.F.-A., and S.G.M.), designed for self-administration, pretested with 10 practicing pediatricians, and revised before it was distributed. There were 31 questions, and it took an average of 10 minutes to complete.

The survey packets included a cover letter with contact telephone numbers, letters of support from professional organizations, the questionnaire, a self-addressed return envelope, and a $5 bill. Mailings began in November 1997, followed by a second mailing in January 1998. We attempted to call or fax messages to physicians' offices to confirm receipt of the packet only after the first mailing. All procedures were approved by the University of California, San Francisco, Committee on Human Research.

Data analyses were managed using SAS statistical software.¹⁷ Univariate analyses compared proportions or categorical data by the χ² test and continuous variables by the t test. Categories of response were collapsed for comparison purposes. A multivariate logistic regression model was constructed to compare pediatricians with family physicians on their responses to the questionnaire and to compare all physician respondents by type of practice (private vs other). Responses to questions on use of cessation techniques with parents were dichotomized to 50% or more of the time they were implemented compared with less than 50%. Responses to the questions on perceived barriers to the delivery of smoking cessation services to parents were dichotomized to "agree a lot" and "agree some" compared with "agree a little" and "agree not at all." The outcome variables were as follows: asking parents about smoking in the house, spending an average of 3 minutes or more counseling a smoking parent to quit, examining 14 behaviors regarding the management of parents who smoke, and determining the level of agreement with 9 perceived barriers to implement smoking cessation services. The main predictor variable was physician specialty (pediatrician or family physician). The multivariate model adjusted for physician sex, age (continuous), ethnicity (white vs other), birthplace (United States vs other), percentage of patients seen who were aged 5 years or younger (continuous), percentage of work time spent in ambulatory care settings (continuous), year graduated from medical school (before 1970 vs after), location of medical school (United States vs other), type of practice setting (private, academic, community clinic, or managed care), and smoking status (ever vs never).

From the original sample of 1000 physicians, 101 surveys were returned and were determined to be ineligible because of current practice (retired, no ambulatory care, or no children seen), specialty criteria, or an undeliverable address. From the sample of 899 remaining and presumed eligible respondents, we received 499 completed questionnaires, for a response rate of 56%. Comparison of respondents with nonrespondents among presumed eligible participants showed no significant differences by physician sex (P = .32), specialty (P = .26), ethnicity (P = .30), and age (P = .53).

In Table 1, the demographic and practice characteristics of the sample are shown. A higher percentage of pediatricians compared with family physicians were women and born outside the United States. No significant differences were found in the age distribution, proportion of former smokers, and distribution of the sample by type of practice. As expected, pediatricians reported a significantly higher percentage of patients seen who were younger than 5 years, but there were no differences in the amount of time reported spent in ambulatory practice. Only 12 respondents were current smokers.

Table 2 shows responses to questions about smoking to parents of children aged 5 years and younger. Nearly two thirds of pediatricians and family physicians indicated asking parents about smoking. Parents who smoked were queried about smoking in the home more often by pediatricians than by family physicians. However, only about half of the parents were asked about smoking in front of their child in other situations (eg, in the car) by pediatricians or family physicians. Less than half the time did physicians ask about a nonparental adult smoking in any room within the home.

A significantly (P<.01) higher proportion of family physicians reported spending 3 minutes or more counseling a smoking parent to quit, but more than half of all physicians reported spending an average of only 1 to 2 minutes counseling (Figure 1). About 1 of 6 pediatricians and 1 of 11 family physicians reported spending no time at all counseling parents to quit smoking.

Table 3 shows the physician-reported delivery of smoking cessation interventions for smoking parents of young children compared by specialty. Pediatricians were significantly more likely to record parental smoking in the medical record as a problem for the child and discuss a child's health risk from secondhand smoke. Physicians from both specialties reported similar rates of providing general advice to quit, increasing motivation to quit, counseling parents on the health risks of smoking, telling parents that smoking is a bad example for their children, and making a referral back to the personal physician. Family physicians were significantly more likely than pediatricians to have reported using 7 of the 14 practices at least half the time. Four of the 14 practices have been central components in randomized trial studies that have demonstrated increased abstinence among smoking patients when clinicians used these practices. These effective practices included setting a quit date, prescribing nicotine replacement therapy, scheduling a follow-up visit or telephone call, and providing nurse-mediated counseling in the office. Family physicians and pediatricians reported using these 4 strategies less than half the time, even though pediatricians were significantly less likely to have reported using each of these. Referral to a smoking cessation program and providing self-help cessation materials were reported to occur less than half the time by physicians from both specialties, but pediatricians were significantly less likely to have reported these practices. There were no differences by specialty in the practice of referring to the parent's personal physician.

Responses to possible barriers to delivery of smoking cessation services to smoking parents of young children are compared by specialty in Table 4. Most respondents did not agree a lot or some with any of the possible barriers. Compared with family physicians, a significantly higher percentage of pediatricians perceived that counseling parents to quit smoking was not effective because parents would ignore the advice or lacked interest in quitting or because the pediatrician lacked smoking cessation counseling skills. Only 20% to 24% of physicians considered parental anger or adequate reimbursement to be barriers, but nearly twice as many agreed that time would be a concern.

Multivariate logistic regression models compared family physicians with pediatricians in their reported use of smoking cessation practices with parents at least half the time during a routine office visit. The comparisons were adjusted for potential confounding variables (Table 5). Results of the models showed that pediatricians were significantly less likely to have reported use of 5 of the 14 practices that promote smoking cessation in parents at least half the time (Table 5). Pediatricians were significantly less likely than family physicians to report spending 3 minutes or more counseling smoking parents to quit. Pediatricians were also significantly less likely to indicate use of the effective practices, such as referring to a cessation program, asking about willingness to set a quit date, scheduling a follow-up visit, or recommending or prescribing nicotine replacement therapy.

In the same multivariate logistic regression model, other predictors were associated with the outcomes of physician use of smoking cessation techniques. Physicians in private practice, compared with those in all other settings, were less likely to report referring patients to a smoking cessation program (odds ratio [OR], 0.53; 95% confidence interval [CI], 0.35-0.80) or having nursing staff counsel a smoking parent on cessation (OR, 0.48; 95% CI, 0.26-0.89). Older age among physicians was associated with decreased likelihood to ask parents if they smoke in front of the child outside the home, advising parents to quit smoking, and motivating parents to quit. Physicians born in the United States (OR, 0.50; 95% CI, 0.27-0.93) were less likely to report counseling smoking parents for 3 minutes or more. Physicians who graduated from US medical schools were less likely to report referring patients to their nursing staff for cessation counseling (OR, 0.26; 95% CI, 0.12-0.59) and giving cessation pamphlets (OR, 0.54; 95% CI, 0.29-0.96).

Comparison of pediatricians and family physicians by agreement with the 9 perceived barriers showed that only 1 was statistically significant (P<.001). Pediatricians were more likely to agree a lot or some that asking parents about tobacco use would "not be effective since I lack smoking cessation skills" (OR, 3.12; 95% CI, 1.55-6.30). Physicians in private practice, compared with those in all other settings, were more likely to agree that asking about tobacco use would be time-consuming (OR, 1.96; 95% CI, 1.12-3.43) and not effective because parents lack interest in quitting smoking (OR, 1.81; 95% CI, 1.01-3.23).

This survey compares reported behavior among pediatricians and family physicians toward assessment of secondhand smoke, counseling parents who smoke, and perceived barriers to delivery of smoking cessation interventions. Pediatricians were significantly less likely than family physicians to have reported spending 3 or more minutes counseling a smoking parent, referring a parent to a smoking cessation program, asking for a quit date from the parent, scheduling a follow-up visit to discuss smoking cessation, increasing motivation for a parent to quit, and prescribing nicotine replacement therapy. The odds of pediatricians compared with family physicians reporting these practices varied from 0.14 for prescribing nicotine replacement therapy and scheduling a follow-up visit to 0.45 for referring a parent to a smoking cessation program. Furthermore, pediatricians clearly perceived a lack of smoking cessation counseling skills as a barrier to even asking parents about tobacco use. These findings identify a potential gap for pediatricians in addressing secondhand smoke exposure for young children given that the principal contributor in most homes relates to parental smoking.

Our study found that family physicians were more likely to report routine counseling of smoking parents of young children. A possible explanation for these findings is that family physicians see adults and as a result have received more training in smoking cessation counseling. The fact that these family physicians would also be the primary care physician for the parents means that they should respond to the questions as they would for their patients. However, one third of the time these family physicians would refer a smoking parent to his or her personal physician, indicating that the respondent was not the primary physician. Furthermore, based on the physicians' reports, only 40% of the time did parents receive smoking cessation pamphlets from family physicians. We would have expected this percentage to be much higher if the parents were the respondent's patients. Thus, the specialty differences observed are not entirely explained by the likelihood that family physicians were also the parent's physician.

We also found that physicians in private practice settings were less likely to report referring parents to a smoking cessation program or asking nursing staff to counsel on cessation. The development and implementation of educational interventions for clinicians to decrease secondhand smoke exposure among young children will require attention to the needs and reality of physicians in private practice.

Overall, most physicians in this study reported asking questions about parental smoking in the home, discussing risks of secondhand smoke, counseling parents on the health risks of smoking, and trying to motivate parents to quit. However, less than half indicated that they had referred parents to a cessation program, given a pamphlet on smoking cessation, asked about willingness to set a quit date, scheduled follow-up visits, or recommended use of nicotine replacement therapy in their efforts to help a smoking parent quit. These data reflect a lack of physician participation in smoking cessation activities and are consistent with other data on the frequency of cessation counseling of adults in the ambulatory setting.

Since less than half of adults are advised to quit by their own primary physicians, pediatricians may have the opportunity to address this important issue with a smoking adult during pediatric care. The National Cancer Institute Physicians' Guide to Smoking Cessation includes a section on smoking prevention, and this was cosponsored by the American Pediatrics Association. The American Pediatrics Association has made efforts at disseminating these skills through trainers, but a formal evaluation has not been conducted. Pediatricians have not been targeted for tobacco control physician-mediated interventions with a smoking parent of a young child. Despite this, pediatric residents who were taught smoking cessation counseling techniques performed as well as those in other primary care specialties.¹⁸ Data collected from pediatricians and parents indicate that counseling about smoking prevention and cessation and about the harms of secondhand smoke would be welcomed and considered important.^19,20 Barriers to counseling about smoking cessation for parents reported by pediatricians include the perception that techniques are ineffective, feeling ill at ease about giving this advice, lack of time, and fear that parents may feel this is intrusive.^13,15,21 In one survey of parents,²⁰ only 15% reported their smoking was "none of the doctor's business," but more than half believed that talking about smoking was "part of the pediatrician's job."

A recent study²² from Scotland found that an educational intervention with smoking parents of asthmatic children (age range, 2-12 years) had no effect on smoking at 1 year of follow-up. Based on studies in adult medical practices, active involvement of clinicians, use of medical record reminders,²³ and promotion of nicotine replacement therapy as a pharmacological adjunct to smoking cessation counseling²⁴ should be routine practice for physicians in contact with smoking parents in pediatric settings. Thus, given the evidence on the feasibility and effectiveness of training clinicians to assist their patients to quit smoking, pediatric clinicians are uniquely positioned to counsel a smoking parent on cessation.

Two randomized trials to evaluate interventions to decrease secondhand smoke exposure in infants have been conducted. A home-based nurse-delivered intervention was evaluated in 933 infants, of whom 25.2% had smoking mothers.²⁵ Among the 121 infants of smoking mothers who completed the study, infants in the intervention group were exposed to 5.9 fewer cigarettes per day and had a lower prevalence of persistent respiratory tract symptoms even though there was no difference in the urinary cotinine level between infants in the intervention and control groups.²⁵ The second study²⁶ randomized 49 pediatric practices to receive a 45-minute training intervention on brief counseling for new mothers on the first 4 well-child visits. The intervention intended to decrease secondhand smoke and was compared with a hospital packet containing written information received by all mothers. Pediatricians were taught to deliver a 2-minute intervention with an emphasis on counseling smoking mothers (42.5% of the study sample) to quit. At 6 months of follow-up, smoking mothers in the intervention group had higher quit rates (5.9% vs 2.7%) and lower relapse rates (45% vs 55%), providing evidence that a brief intervention can have a positive impact on maternal smoking.²⁶

We found that most physicians in our study did not agree with any of the 8 barriers to delivery of smoking cessation services. Pediatricians were less convinced that parents would be interested in quitting and more concerned about parents ignoring their advice. However, 26% agreed that they lacked smoking cessation counseling skills, compared with only 7% of family physicians. Based on these findings, training pediatric clinicians to assist smoking parents to quit seems feasible and should become a priority for residency programs.

Our study has important limitations. First, we obtained a response rate of 56%, which may introduce bias even though a comparison of available demographics did not show significant differences between respondents and nonrespondents. Moreover, this level of response rate is comparable to other physician surveys.²⁷ Second, all the responses are self-reported behavior or estimates and none were validated with intermediate outcomes or surveys of patients' parents. However, physicians are more likely to overestimate the frequency of prevention activities in practice and thus our data probably represent a best-case scenario. An intervention study should incorporate a validation component to evaluate physician self-report. Finally, these physicians reside in California, where there has been an active antismoking campaign for about 10 years, and their sensitivity and responses to these questions may be affected by the general change in societal norms on tobacco.

In conclusion, our study confirms that pediatricians and family physicians are not taking full advantage of an important opportunity to decrease secondhand smoke exposure in their young patients. Training pediatricians and family physicians in smoking cessation counseling techniques and motivating them to use these with smoking parents may help reduce secondhand smoke exposure among children. Enhancing the practice environment to facilitate smoking cessation counseling of parents of young children, increasing referrals to smoking cessation programs, and teaching all clinicians how to recommend nicotine replacement therapy and other pharmacological adjunctive therapy for nicotine addiction are additional practices that may reduce secondhand smoke exposure for children.

Accepted for publication August 22, 2000.

This study was supported by grant 6RT-0368, pediatrician training to prevent children's tobacco exposure, from the University of California Tobacco-Related Disease Research Program.

Corresponding author: Eliseo J. Pérez-Stable, MD, Division of General Internal Medicine, University of California, San Francisco, 400 Parnassus Ave, Room A-405, Box 0320, San Francisco, CA 94143-0320 (e-mail: eliseops@medicine.ucsf.edu).