Abstract
BACKGROUND: General practitioners (GPs) often feel uncomfortable when patients request an antibiotic when there is likely little benefit. This study evaluates the effect of access to point-of-care tests on decreasing the prescription of antibiotics in respiratory tract infections in subjects who explicitly requested an antibiotic prescription.
METHODS: Spanish GPs registered all cases of respiratory tract infections over a 3-week period before and after an intervention undertaken in 2008 and 2009. Patients with acute sinusitis, pneumonia, and exacerbations of COPD were excluded. Two types of interventions were performed: the full intervention group received prescriber feedback with discussion of the results of the first registry, courses for GPs, guidelines, patient information leaflets, workshops, and access to point-of-care tests (rapid streptococcal antigen detection test and C-reactive protein test); and the partial intervention group underwent all of the above interventions except for the workshop and access to point-of-care tests.
RESULTS: A total of 210 GPs were assigned to the full intervention group and 71 to the partial intervention group. A total of 25,479 subjects with respiratory tract infections were included, of whom 344 (1.4%) requested antibiotic prescribing. Antibiotics were more frequently prescribed to subjects requesting them compared with those who did not (49.1% vs 18.5%, P < .001). In the group of GPs assigned to the partial intervention group, 53.1% of subjects requesting antibiotics received a prescription before and 60% after the intervention, without statistical differences being observed. In the group of GPs assigned to the full intervention group, the percentages were 55.1% and 36.2%, respectively, with a difference of 18.9% (95% CI: 6.4%–30.6%, P < .05).
CONCLUSIONS: Access to point-of-care tests reduces antibiotic use in subjects who explicitly request an antibiotic prescription.
Introduction
Many patients seek medical attention for respiratory tract infections such as cough, cold, and sore throat, most of which are viral in origin and can be managed without antibiotic therapy. Nonetheless, patients are often prescribed antibiotics for these conditions.1 The overuse and/or misuse of antibiotics can lead to significant consequences, including increased cost, bacterial resistance, therapeutic failure, and adverse effects including drug toxicities and drug interactions.2,3 The perception of general practitioners (GPs) regarding patient expectations for a prescription is a strong predictor for antibiotic prescribing.4 When patients expect antibiotics, they are more likely to be prescribed,5 and likewise, when physicians perceive that patients expect antibiotics.6–8 Different studies show that GPs have difficulties in determining whether the patients actually expect antibiotic therapy.9–11 There is compelling evidence that patients' satisfaction with the consultation is not affected by prescribing of antibiotics, and patient dissatisfaction has been shown to be significantly related only to poor communication between the patient and the doctor.12,13
Despite numerous papers on patients' expectations and physicians' perceptions and their impact on antibiotic prescribing for respiratory tract infections, few studies addressing the explicit request for antibiotics by patients during the consultation have been published. GPs often prescribe an antibiotic to fulfill patient demands in an attempt to satisfy the patient. GPs very often feel uncomfortable when coping with this demand, making the prescription of an antibiotic more likely compared with situations in which this demand does not occur.14
A prospective non-randomized before-and-after study was performed in primary care clinics in Spain as a part of the Health Alliance for Prudent Prescribing, Yield and Use of Antimicrobial Drugs in the Treatment of Respiratory Tract Infections (HAPPY AUDIT) project, a study financed by the European Commission.15 The main objective of this study was to strengthen the surveillance of respiratory tract infections in primary health care through the development of intervention programs targeting GPs and patients. GPs from 6 countries participated in this study (Denmark, Sweden, Lithuania, Russia, Spain, and Argentina). The aim of the present study was to investigate predictors for subjects' request of antibiotics and evaluate the effect of access to point-of-care tests on antibiotic prescribing for respiratory tract infections among patients who explicitly requested a prescription of antibiotics.
QUICK LOOK
Current knowledge
Indiscriminate antibiotic prescribing practices can lead to ineffective therapy and may promote antibiotic resistance. Physicians are frequently pressured by patients with symptoms for an antibiotic prescription in the absence of confirmed bacterial infection. This conflict disadvantages both the physician and the patient.
What this paper contributes to our knowledge
Access to point-of-care testing results to confirm that infection reduced antibiotic use in subjects who explicitly requested a prescription for antibiotics. Availability of point-of-care testing results reduces unnecessary antibiotic utilization. Access to these results reduced antibiotic use in subjects who explicitly requested a prescription for antibiotics.
Methods
GPs were selected on a voluntary basis and registered all subjects with respiratory tract infections during a 3-week period, covering a total of 15 work days, in the winter months before (pre-intervention) and after an intervention (post-intervention). The data were registered on a sheet following a prospective self-registry methodology. The physician noted specific parameters related to the medical care, including subject age and sex, symptoms and signs, use of point of care tests (rapid streptococcal antigen detection test and C-reactive protein test), assumed diagnosis, treatment (decision and choice of antibiotics), and whether the subject requested an antibiotic.15 Common cold, otitis media, pharyngitis, acute bronchitis, influenza, and other suspected viral respiratory tract infections were considered in this study. Patients with exacerbations of chronic bronchitis and/or COPD and pneumonia were excluded. Ethical approval was obtained from the Fundació d'Investigació d'Atenció Primària Jordi Gol (Barcelona, Spain; reference 44154).
Detailed information about the intervention can be found in the study protocol.16 Briefly, 2 types of interventions were undertaken. One intervention (partial intervention) consisted of meetings with the GPs including prescriber feedback based on the results from the first registration, training courses on the diagnosis and treatment of respiratory tract infections, and review of guidelines on respiratory tract infections. The other intervention (full intervention) also involved access to 2 point-of-care tests in practice: (1) rapid antigen detection tests for the diagnosis of streptococcal pharyngitis and (2) C-reactive protein tests to support the GPs in distinguishing suspected bacterial from viral infections. GPs assigned to the full intervention group were instructed not to use the point-of-care tests as stand-alone tests but rather as additional tests in case of doubt: using rapid antigen detection test in subjects with pharyngitis and at least 2 Centor criteria (temperature > 38°C, tonsillar exudate or inflammation, tender cervical glands, and absence of cough),17 and the C-reactive protein test in severe lower respiratory tract infections.18 They were advised to withhold antibiotics in subjects with negative rapid antigen detection test results and C-reactive protein values < 20 mg/L and to consider antibiotic prescription in subjects with a positive rapid antigen detection test and/or a C-reactive protein result of > 100 mg/L.
Statistical Analysis
Descriptive and bivariate analyses were conducted. To identify potential predictors for subject request of antibiotics, we performed a multivariate logistic regression analysis that included age, prior duration of symptoms, gender, signs, and symptoms. Variables with P < .1 in the bivariate analysis were selected for inclusion in the multivariate model, and the final selection of variables was performed by the backward stepwise selection analysis. Significant differences were considered with a P value < 0.05.
Results
Figure 1 describes the general scheme of the study. A total of 25,479 subjects with respiratory tract infections were included, and, according to the GPs, 344 (1.4%, 95% CI [CI] 1.3–1.6) asked for an antibiotic, mainly for lower respiratory tract infections (Table 1). More subjects requested antibiotics from GPs assigned to the full intervention than from GPs allocated to the partial intervention (1.5% vs 1%, P < .05) (Fig. 1). Among subjects requesting antibiotics (n = 344), 169 (49.1%) received an antibiotic. Antibiotics were prescribed significantly less frequently when subjects did not request an antibiotic (18.5%, P < .001). The difference was, however, more obvious in those with a common cold, influenza, pharyngitis, and acute bronchitis (Fig. 2).
In the group of GPs assigned to the partial intervention (without point-of-care tests), 53.1% (95% CI: 35.8%–70.4%) of subjects were prescribed antibiotics before the intervention and 60% (95% CI: 44.8%–75.2%) after the intervention. In the group of physicians assigned to the full intervention (with access to rapid tests), 55.1% (95% CI: 47.3%–62.9%) of subjects were prescribed antibiotics before the intervention and 36.2% (95% CI: 27.4%–44.9%) after the intervention, with a difference of 18.9% (95% CI: 6.4%–30.6%, P < .05) (Fig. 3).
Table 2 shows the results of the multivariate analyses of the different predictors for subjects' request of antibiotics. Only age, fever, pain when swallowing, increased sputum production, and purulent sputum were significantly associated with a demand for antibiotics.
Of the 344 antibiotic-requesting subjects, rapid antigen detection tests were performed in 41 (11.9%) and C-reactive protein in 12 (3.5%). All point-of-care tests were undertaken by GPs assigned to the full intervention after the intervention. The streptococcal rapid test was negative in 35 subjects, with antibiotics being prescribed in only one case, whereas they were prescribed in the 6 positive results. Of the 332 subjects requesting antibiotics in whom a C-reactive protein test was not performed, antibiotics were given to 49.7% of the subjects, whereas they were only given in 4 cases out of the 12 subjects in whom this rapid test was performed (33.3%).
Discussion
This study shows that antibiotics are more likely to be prescribed to subjects with acute respiratory tract infections who request them than to subjects who do not explicitly request antibiotics. Moreover, an intervention aimed at promoting more prudent use of antibiotics by GPs can reduce the prescription of antibiotics, mainly when the GPs have access to rapid tests in their consultations. Negative rapid antigen detection test results and very low C-reactive protein values were associated with a lower prescription rate among antibiotic-requesting patients.
This study has some limitations. This study was not a clinical trial, and the groups were not assigned randomly. The percentage of physicians who stated that subjects had requested an antibiotic was very low in our study (only 1.4%), being much lower than what was reported by Coenen et al,13 who observed a percentage of 10.2% in a study involving nearly 3,500 adult subjects with acute cough in 14 different European networks. In our study, approximately one third of the infections included corresponded to the common cold; in these cases, the request for an antibiotic was unlikely. However, subjects more frequently requested antibiotics for lower respiratory tract infections. In our study, GPs were asked to tick off the box of request only when subjects explicitly demanded an antibiotic during the consultation for respiratory tract infection. Participation in a study on the rational use of antibiotics may have also influenced the GPs to prescribe antibiotics more rationally; however, the same GPs registered both the first and second registries.
To our knowledge, no other study has examined the effect of point-of-care tests on antibiotic prescribing when subjects request an antibiotic. In our study, GPs assigned to the partial intervention group, that is, without access to rapid tests, failed to reduce the prescription of antibiotics for antibiotic-requesting subjects, despite having undertaken a multifaceted intervention including feedback with discussion of their own results, training in guidelines, and the use of leaflets for subjects to back up their decision. However, training in communication skills, which has been shown to effectively deal with patients without prescribing antibiotics in other studies, was not carried out in the HAPPY AUDIT study.19–21 Only those GPs assigned to the full intervention group significantly reduced the amount of antibiotics prescribed after the intervention had taken place. The only difference between the full and the partial intervention groups was the access to rapid tests, and this was associated with a 18.9% reduction in antibiotic prescribing. GPs might have been less likely to be influenced by subjects' demand for antibiotics when they had access to a rapid test that could help them to convince subjects that they did not need to take antibiotics.22 This statement is also supported by the results of a published qualitative study aimed at exploring the views and experiences of 66 GPs using 2 interventions to optimize consultations for respiratory tract infections.23 In this study, the C-reactive protein test was praised by the GPs, as it gave additional diagnostic information, which reduced the uncertainty as to whether antibiotics might be of value.23 In these cases, GPs felt that the test supported a non-prescription decision, where relevant, and provided reassurance to subjects.
Conclusions
Patients' expectations for antibiotic prescribing have a strong influence on GPs' prescribing habits. In this audit-based study, GPs were more prone to prescribe an antibiotic when subjects explicitly requested it. This study also shows that physicians with access to rapid diagnostic tests prescribed fewer antibiotics when subjects explicitly asked for them in the consultation.
Acknowledgments
We thank all the primary care physicians who participated in this study.
HAPPY AUDIT Investigators
Andalusia, Spain: Juan de Dios Alcántara, Carolina Álvarez, Francisco Atienza, Manuel Baeza, Juan Bellón, Inmaculada Carrillo, César J Costa, Pedro Crespo, Carmen Dastis, Salvador Domínguez, M Magdalena Gálvez, M Isabel González, Aurora Guerrero, Carmen Gutiérrez, Rosa del Pilar Herrera, Guillermo Largaespada, Beatriz López, Inés M López, María Luisa Manzanares, Leonor Marín, Francisco Marmesat, M Mercedes Martínez, Rocío Martínez, M Inmaculada Mesa, Yolanda de Mesa, Guillermo M Moreno, M Luisa Moya, José Oropesa, Carolina Pérez, Manuel Pérez-Cerezal, Juan J Quero, M Pilar Rojo, Miguel Sagristá, Consuelo Sampedro, Jesús Carlos Sánchez, José Cristobal Sendín, Miguel Silva, José Miguel Solís, Laura Suárez, Irene Victoria de Tena, Salvador Torres.
Asturias, Spain: Carmen Alonso, Fernando Álvarez, M Etelvina Castañón, Beatriz Fernández, Guillermo García, M Pilar Jimeno, Ramón Macía, Carmen Martín, Ernesto Martínez, Covadonga Monte, M Amor Paredes, Javier Pérez, M Mar Pizarro, Felipe J Rodríguez, Celia Teresa Tamargo, Salvador Tranche, M Raimunda Vázquez.
Balearic Islands, Spain: Ester Adelantado, Antonio Ballester, Arnest Bordoy, Bernardino Bou, José Antonio Chumillas, Francisca Comas, Teresa Corredor, Consuelo Corrionero, Esther Domínguez, Alberto Eek, Teresa Estrades, Sebastián Fluxa, Maria E Garau, Josefa García, Yolanda Garzón, M José González-Bals, Rosa Grimalt, Antoni J Jover, Catalina Llabrés, Magdalena Llinas, Marian Llorente, Montse Llort, Ana M Macho, David Medina, Susana Munuera, Joana M Oliver, Rosa Ortuño, Juana Pérez, Lourdes Quintana, María Martín-Rabadán, José Alfonso Ramón, Jaime Ripoll, José Ramón Roca, Raquel Ruano, M Carmen Santos, Isabel M Socias, M Mar Sureda, Carolina Tomás.
Canary Islands, Spain: Margarita Aguado, Pilar Aguilar, M Carmen Artiles, Jonás de la Cruz Cabrera, M Isabel Cardenes, Encarna Duque, José Luis Eguren, Javier Francés, Alicia González, Gloria Guerra, M de las Huertas Llamas, Alicia Monzón, Aurelia Perdomo, Carlos Prieto, José Luis Rodríguez, M Rafaela Sánchez, M Teresa Simó, Lucía Tejera, José M Toscano.
Catalonia, Spain: Maite Aizpin, Francesc Arasa, Josep Ausensi, Teresa Aviñó, Àngel Ayala, Montse Balagué, Jaume Banqué, Jaume Baradad, Marta Beltrán, Josep Lluis Berdonces, Noemí Bermúdez, Armando Biendicho, Javier Blasco, Miriam Boira, Enric Buera, Joan Cabratosa, Sonia Castro, Mireia Cid, Maria de Ciurana, M Rosa Dalmau, Carmen Delgado, Teresa Escartín, Rosario Espinosa, Gemma Estrada, Eugeni Fau, M Emma Fava, Ester Fernández, María Ferré, Pilar Franco, Joaquim Franquesa, Elena Esther Fuentes, Carme García, Manuel García, Montse García, Sergi Giró, Mònica Gómez, Yosbel Guerra, Silvia Hernández, Francisco Ibáñez, Roland Juan, Josep Lluis Llor, Ana Luque, Anna Manzanares, Emili Marco, Judith Marin, Emma Marqués, Ignacio Martínez, Maribel Martínez, Rosa Martínez, Ariadna Mas, Cinta Monclús, Pau Montoya, M Luisa Morató, Jesús V Muniesa, Esther Mur, M Assumpció Nadal, Elena Navarro, Miquel Navarro, Carme Pascual, Marina Payà, Almudena Pérez, Pilar Pérez, Cristina Pozo, Luis Quinzaños, Anna Ràfols, Mercè Ribot, Maria Riera, Pilar Rivera, Carolina Robado, Purificación Robles, M José Roig, M Carmen Ros, José Miguel Royo, M Victoria Rubio, Anna Serra, M Ángeles Sieira, Yaiza Sierra, Hiam Tarabishi, Silvia Torrent, Leticia Troyano, María Úbeda, Antonio Ubieto, Susana Vargas, Jordi Vilano, Assumpció Wilke.
Galicia, Spain: Eduardo Alonso, Margarita Bayón, Alejandro Cardalda, Francisco Castrillo, Ángeles Charle, Marina Cid, Pilar Cobas, Peregrina Eiroa, Ana Fernández, Elena García, Ricardo Manuel Héctor, Susana Hernáiz, Jesús De Juan, Pilar Mendos, Elisa Mosquera, Concepción Nogueiras, Ana M Ogando, Elías Puga, Adolfo Rodríguez, José Benito Rodríguez, Coro Sánchez, Joaquín San José, Santiago Santidrián, Luis Seoane, M Concepción Soutelo, Jesús Sueiro.
Madrid, Spain: Ana Aliaga, Raquel Baños, José M Casanova, Santiago Castellanos, José Corral, Angélica Fajardo, Antonia García, M Begoña García, Alfredo Gómez de Cádiz, Manuel Gómez, Paloma González, M Teresa Hernández, Paloma Hernández, M del Canto de Hoyos, M Carmen López, Rosa Martín, Isabel Miguel, José M Molero, Joaquín Morera, Alicia Muñoz, Francisco Muñoz, Javier Muñoz, M Soledad Núñez, Lourdes Pulido, Ana Rodríguez, Carmen Rodríguez, Ana Ruiz, Ernestina Ruiz, Javier de la Torre, Amalia Velázquez, Lourdes Visedo, Antonio Yagüe.
Rioja, Spain: Carmen Babace, M Mar Díez, Francisco García, Tomas García, Jesús Ortega, Olga Revilla, Rosa Ruiz de Austri, José Luis Torres, Antonio Trueba, Santiago Vera.
Valencia, Spain: Dolores Alfonso, Manuel Batalla, Lourdes Bermejo, Nuria Bosch, Cristina Calvo, Beatriz Camaro, M Pilar Carceller, Manuela Domingo, Manuel Galindo, Carmen Gandía, Concepción García, Andrés Vicente Genovés, M Carmen González, M Ángeles Goterris, Amparo Haya, Eva M Izquierdo, Carmen Lloret, Engracia López, Pilar Marín, M José Martín, Susana Martí
Footnotes
- Correspondence: Carl Llor MD PhD, Department of Primary Care and Public Health, School of Medicine, Cardiff University, 5th Floor, Neuadd Meirionnydd, Heath Park, Cardiff CF14 4XN, United Kingdom. E-mail: llorc{at}cardiff.ac.uk.
This study is a Specific Targeted Research Project (STREP) funded by The European Commission: DG SANCO under the Frame Program 6 (SP5A-CT-2007-044154, contract 044154).
Dr Llor received a grant from the Fundació Jordi Gol i Gurina for a research stage at the University of Cardiff, as well as research grants from the European Commission (Sixth and Seventh Programme Frameworks), Catalan Society of Family Medicine, and Instituto de Salud Carlos III (Spanish Ministry of Health). The other authors have disclosed no conflicts of interest.
- Copyright © 2014 by Daedalus Enterprises