Cost-Effectiveness of a New Outpatient Pulmonology Care Model Based on Physician-to-Physician Electronic Consultation

Introduction This study assesses the impact of an electronic physician-to-physician consultation program on the waiting list and the costs of a Pulmonology Unit. Materials and Methods A prepost intervention study was conducted after a new ambulatory pulmonary care protocol was implemented and the capacity of the unit was adopted. In the new model, physicians at all levels of healthcare send electronic consultations to specialists. Results In the preintervention year (2019), the Unit of Pulmonology attended 7,055 consultations (466 e-consultations and 6,589 first face-to-face visits), which decreased to 6,157 (3,934 e-consultations and 2,223 first face-to-face visits; 12.7% reduction) in the postintervention year (all were e-consultations). The mean wait time for the first appointment was 25.7 days in 2019 versus 3.2 days in 2021 (p < 0.001). In total, 43.5% of cases were solved via physician-to-physiciane-consultation. A total of 2,223 patients needed a face-to-face visit, with a mean wait time of 7.5 days. The mean of patients in the waiting listing decreased from 450.8 in 2019 to 44.8 in 2021 (90% reduction). The annual time devoted to e-consultations and first face-to-face visits following an e-consultation diminished significantly after the intervention (1,724 hours versus 2,312.8; 25.4% reduction). Each query solved via e-consultation represented a saving of €652.8, resulting in a total annual saving of €827,062. Conclusions Physician-to-physiciane-consultations reduce waiting times, improve access of complex patients to specialty care, and ensure that cases are managed at the appropriate level. E-consultation reduces costs, which benefits both, society and the healthcare system.


Introduction
A public health system must be built on the principles of equity by providing services to all patients irrespective of their economic or social status; efectiveness, by solving health problems at the adequate level of healthcare and preventing unnecessary medical procedures and unneeded complementary tests studies; quality, by using quality indicators to assess clinical activity and health outcomes and identify dysfunctions; and transparency, by making results public [1,2]. To attain these goals, it is necessary to exploit information systems and promote the implementation of collaborative networks at the diferent levels of care [3], which will ultimately guarantee the sustainability of the system [4].
Waiting lists are one of the most important challenges that prevent our public health system from complying with these principles. Before the intervention, the volume of outpatient consultations had increased in our third-level hospital Unit of Pulmonology. Tis increase was partially due to the interventions implemented to reduce the number of hospital admissions (one-stop consultations, day hospital, and outpatient management of patients undergoing invasive diagnostic procedures, to name a few). Although waiting lists are unavoidable, long waitlists may compromise the principle of equity and raise ethical concerns about distributive justice, inequalities, or discrimination [5]. Telemedicine based on electronic physician-to-physician consultations (e-consultations) may help reduce waiting lists, thereby improving healthcare equity, efectiveness, and clinical outcomes [6]. Trough e-consultations, physicians provide clinical information and send consultations securely to specialists (in this case, a pulmonologist), via the electronic health record of patients with good reported results in other specialties [6,7]. By this method, the specialist decides whether a patient needs a face-to-face appointment with specialty care, or otherwise the case can be managed in primary healthcare (PH) following the specialist's advice.
Te purpose of this study was to assess the efects of a physician-to-physiciane-consultation model on wait times and costs associated with patient referrals from primary or specialty care to the Pulmonology Unit of a third-level hospital in Spain.

Materials and Methods
Te study was conducted in a tertiary 1000-bed university hospital serving a population of 450,000 distributed across 46 city councils, with 24.2% of the population being older than 65 years. For the population older than 14, there are 301 general practitioners (GPs) distributed across 56 primary healthcare centers and 21 rural clinics.
In this pre and postintervention study (implementation of a physician-to-physiciane-consultation model), the year 2019 corresponds to the preintervention period, whereas 2021 corresponds to the postintervention period. Te year 2020 was considered as the "washout" period, since e-consultations started in June 2020, the two models overlapped during the frst weeks, and the outbreak of the COVID-19 pandemic would have infuenced the results obtained.
In 2019, the Pulmonology Unit had 15 open slots per weeks for e-consultations and 156 slots per week for frst faceto-face appointments. In 2020, an analysis was performed of the results of 2019 to estimate the resources that were necessary to implement the e-consultation model. Te capacity of the Pulmonology Unit increased to 157 e-consultations and 96 frst face-to-face consultations per week (44 weeks/year). Te estimated mean time devoted to an e-consultation and frst face-to-face visit was 15′ and 20′, respectively (both, for the pre and the postintervention period).
Until 2019, GPs were given the option to choose between a conventional face-to-face visit or an e-consultation. From June 1 st 2020, GPs and specialty physicians could only refer patients to the Unit of Pulmonology via e-consultation. Te established protocol is as follows: an initial telematic evaluation (e-consultation) is performed to sort out patients with mild conditions from patients with severe diseases. Te pulmonologist has to respond to the e-consultation within 3 days. Te pulmonologist refers patients with mild conditions back to their referring physician via another e-consultation indicating a treatment plan (management of stable patients or with minor symptoms).
It was estimated that 40% of cases would be solved via econsultation. If a patient needed a face-to-face visit, the pulmonologist could also order, where appropriate, a chest X-ray, blood test, a repeat lung function test to complement the studies performed in PH (FeNO, CO difusion, and static lung volumes), and/or respiratory polygraphy. For this model to be efective, coordination was required with the Department of Pulmonary Function Testing, the Sleep Unit, and the Department of Radiology. Tus, since these departments were potential bottlenecks of the new model [8], some slots were reserved for basic studies. Te faceto-face visit, along with the complementary studies requested, has to be scheduled for a suitable date based on whether the patient is preferential or not. Face-to-face appointments, including appointments for complementary tests, had to be scheduled within 3 weeks after the econsultation. Ten, the pulmonologist can either refer the patient back to the referring physician or manage the case directly in the Pulmonology Unit. Finally, outpatient care agendas were adapted to the estimated demand based on data from previous years. Figure 1 shows the econsultation protocol. Waiting time (in days) was established according to the date of consultation considered appropriate by the referring physician. In case such a date was not detailed on the electronic request, it was calculated from the day after the date of request.
Costs were estimated based on (i) the ofcial fees published by the Servicio Gallego de Salud [9] (40% lower for e-consultations), multiplied by 0.19 €/Km [10] if the patient was referred to the hospital [11]; (ii) time to travel to the hospital including the duration of the visit (weighted average time to reach the hospital, park the car, reach the pulmonology ward, and time spent in the face-to-face visit (79′)); (iii) costs depending on whether the population was active [based on patient average gross income (€18,768.21) and unemployment rate (7.9%)] [12]; or retired (2.3% of subjects >65 are engaged in volunteer work [13] and 22.6% perform activities that contribute to the gross domestic product) [14]; and (iv) costs related to leisure time (47% of labor cost) [15], wait time (101.5 €/month) [16], and CO 2 emissions per private car travel (volume of CO 2 emitted per kilometer and cost of each CO 2 gram) [17][18][19][20].

Statistical Analysis
Continuous variables are expressed as mean ± standard deviation or as median values [interquartile interval] for asymmetric distributions. X 2 and Mann-Whitney U tests were used to analyze the diferences between consultation models. Data analysis were performed using SPSS version 22.0 (SPSS Inc., USA). A total of 7,055 consultations (466 e-consultations and 6,589 frst face-to-face visits) were performed in the preintervention year (466 e-consultations and 6,589 frst faceto-face visits) ( Figure 2). In the postintervention year, 3,934 e-consultations were conducted (44.2% reduction), which generated 2,223 frst face-to-face visits (12.7% reduction). Figure 3 details the monthly distribution of econsultations and the associated frst face-to-face visits conducted in 2021. On average, there were 327.8 econsultations (range, 256-401) and 185.3 face-to-face visits (range, 107-256) per month. As many as 43.5% of cases were solved via e-consultation. Table 1 describes the clinical-epidemiological characteristics of patients seen in 2019 and 2021. Tere were no statistically signifcant diferences in terms of age, sex, or any of the associated comorbidities.  Table 2 shows the annual time spent by our pulmonologists on e-consultations and face-to-face visits in 2019 and 2021. In the preintervention year, the total time devoted to consultations was 2,312.8 hours. In the postintervention year, although the capacity of the unit had been increased, the total time spent on this type of consultation was 1,724.5 hours (25.4% reduction). Table 3 summarizes the results of the economic study, which reveals that each query that is solved with econsultation results in a saving of €652.8. Considering that 56.5% of patients required a face-to-face appointment and the mean wait time was 7.5 days vs. 25.7 days for conventional consultations, the e-consultation program represented a saving of €827,061.6.

Discussion
Te results obtained suggest that our outpatient pulmonology care model based on an initial evaluation through an e-consultation reduces waiting time signifcantly. Tis model makes it possible to sort out less complex cases that do not require a face-to-face visit; this way, waiting times are reduced for more complex cases that actually need specialty care. As compared to face-to-face visits, e-consultations result in a cost reduction, which benefts both, the healthcare system and society. In early 2020, the situation in the Outpatient Unit of Pulmonology was concerning, as the demand for frst appointments had increased by 125.7% in the last decade, and the mean waiting time for the frst appointment was 25.7 days in the previous year (monthly range: 12.2-47.1). Tis situation was unacceptable for patients and could cause the collapse of outpatient pulmonology services. Although waiting time can be reduced by reinforcing staf, this would not have diminished the sustained demand increase. By only hiring more staf, we would have faced the same problem in a few years. In conventional outpatient care, a case is not actually evaluated until the frst face-to-face visit. If during the frst visit the physician requests complementary studies to establish a diagnosis and/or assess disease severity, the results will not be available until the following visit some weeks later. Tis model delays the fnal diagnosis and initiation of treatment. Tis renders the model clearly inefcient.
According to Oseran and Wasf [21], e-consultations should facilitate physician-to-physician communication that allow asynchronous communication, enable requests and responses to be made via a secure electronic system and recorded on the patient's medical record, and address a specifc health problem. Our e-consultation model meets all criteria.
Te implementation of this model reduced signifcantly the mean waiting time. Although there are previous experiences in other medical specialties, to the best of our knowledge, this is the frst time an e-consultation model is applied to pulmonology [22][23][24][25][26][27]. In total, 43.5% of cases were solved via e-consultation, slightly exceeding our estimates, and in an intermediate range with respect to previous studies (range: 21.4-69%) [6,22,23]. Te proportion of cases solved via e-consultation is expected to decrease over the years as a result of two efects: learning, as clinicians will progressively improve their patient management skills; and dissuasive, due to an increasing awareness that minor respiratory problems are rarely referred to a face-to-face consult. Tis would partially explain the reduction of referrals observed after the implementation of the econsultation model. If our estimates are correct, the number of annual e-consultations will decrease, as it already occurred during the frst year of intervention. In addition, the number of face-to-face appointments following an econsultation should remain stable, since referral criteria will not presumably change over time. Tis way, demand for outpatient care will progressively stabilize, which will make it easy to estimate the annual demand for face-to-face appointments and calculate the number of resources required for this type of consultation.
Our results suggest that e-consultations prevent unnecessary actions, foster the management of health problems at the appropriate level of healthcare, and possibly reduce costs by avoiding unnecessary medical procedures, complementary studies, and travel [21,24,25]. Tis contributes to ensuring the sustainability of the healthcare system. In addition, this model facilitates the provision of timely specialty care to patients with severe conditions and optimizes the use of resources. Te reduction in the mean wait time achieved in our study is consistent with that reported in previous studies, although they were performed in a Unit of Cardiology [26]. In a similar study, Winchester et al. [26] achieved to reduce the wait time for the frst appointment from 24 to 13 days (46%) and estimated that 60% of cases could be managed via e-consultation since they were related to patients with stable disease or mild symptoms. Comparison of results is not possible since, to the best of our knowledge, there are no previous experiences in other units of pulmonology in the literature. Of the 21 studies analyzed in a systematic review, 6 involved patients with heart problems, whereas the remaining 15 involved patients of diferent specialties [21]. Te reduction in waiting time could be partially attributed to the increased capacity of the Unit of Pulmonology. Tis is theoretically true, although the total time spent on consultations was signifcantly lower in the postintervention year, as compared to the preintervention year (1,724.5 hours vs. 2,312.8, respectively; 25.4% reduction). Tis can be explained by the fact that the total number of consultations was lower in the postintervention year and dramatically below the number of consultations expected for 2021 (3,109 hours). Nevertheless, if the number of consultations had grown in 2021, the mean wait time to an econsultation would not have increased, given that the capacity of the unit was increased; hence, the unit would still have been able to meet the demand for e-consultations.
Concerns have been raised that, although econsultations reduce wait time, the rate of complications may rise. However, this concern is not supported by the results reported in the literature. In a randomized clinical trial conducted in a Unit of Cardiology, Olayiwola et al. observed that the number of admissions to the Emergency Department was lower in the intervention group [23]. In another experience in a Unit of Cardiology, Wasfy et al. confrmed that the e-consultation model is safe and efective and improves the efcacy of outpatient care [27]. Finally, Rey-Aldana et al. implemented an e-consultation model in a Unit of Cardiology in our healthcare district and concluded that this model is safe since it improved waiting time        [6]. In our case, there are no data available in the literature about health outcomes in terms of ED admissions, hospitalizations, and mortality, which we will investigate in future studies. Considering the results obtained, the use of econsultations for the management of patients referred from PH to a Unit of Pulmonology is efective in reducing direct and indirect costs incurred by patients and the healthcare system, as compared to the conventional faceto-face visit model. According to a study undertaken to assess the advantages of telemedicine, the cost of an econsultation is $120 vs. $228 for face-to-face visits. However, these authors did not consider some of the costs that we included in our estimations [28]. Paquette and Lin assessed environmental pollutant emissions resulting from patient travels (carbon dioxide, carbon monoxide, nitric oxide, and volatile organic compounds) [29]. However, unlike in our study, the saving resulting from the reduction of emissions was not calculated in that study.
Tis study has some limitations. Te intervention was implemented in a specifc specialty unit, and the impact of this e-consultation model may be diferent in other specialties [30]. Although our consultations are 100% open since January 1 st , 2021 without any restrictions at all, limitations in PH, added to patients' reluctance to go to the hospital due to the COVID-19 pandemic, may have contributed to the reduction in the number of consultations. Finally, the occurrence of complications was out of the scope of this study. Te data provided about this issue are based on the results provided in the literature.
In summary, the results obtained in this study suggest that an initial evaluation via an e-consultation in outpatient pulmonology care reduces signifcantly the waiting time and facilitates the identifcation of less complex cases that do not require a face-to-face visit, whereas it improves access of complex patients to specialty care. Tis model ensures that clinical problems are managed at the appropriate level of healthcare and contributes to the reduction of healthcare costs. Te implementation of this model would improve the efcacy of outpatient pulmonology care.
Abbreviations e-consultation: Electronic consultation ED: Emergency department €: Euros GP: General practitioner PH: Primary healthcare.

Data Availability
Te data used to support the fndings of this study were supplied by Luis Valdés under license and so cannot be made freely available. Requests for access to these data should be made to Luis Valdés (Luis.valdes.cuadrado@ sergas.es).

Conflicts of Interest
Te authors declare that they have no conficts of interest.

Authors' Contributions
José ManuelÁlvarez-Dobaño conceived and designed the study, performed data analysis and interpretation, reviewed intellectual content, and approved fnal manuscript. M a Elena Toubes conceived and designed the study, performed data analysis and interpretation, reviewed intellectual content, and approved fnal manuscript. JoséÁngel Novo-Platas carried out data acquisition, reviewed intellectual content, and approved fnal manuscript. Francisco Reyes-Santías performed the analysis and interpretation of the data, performed economic analysis, reviewed intellectual content, and approved fnal manuscript. Gerardo Atienza performed data analysis and interpretation, reviewed intellectual content, and approved fnal manuscript. Manuel Portela performed data analysis and interpretation, reviewed intellectual content, and approved fnal manuscript. Carlos Rábade performed data analysis and interpretation, reviewed intellectual content, and approved fnal manuscript. Tamara Lourido performed data analysis and interpretation, reviewed intellectual content, and approved fnal manuscript. Ana Casal performed data analysis and interpretation, reviewed intellectual content, and approved fnal manuscript. Carlota Rodríguez-García performed data analysis and interpretation, reviewed intellectual content, and approved fnal manuscript. Vanessa Riveiro performed data analysis and interpretation, reviewed intellectual content, and approved fnal manuscript. Romina Abelleira performed data analysis and interpretation, reviewed intellectual content, and approved fnal manuscript. Jorge Ricoy performed data analysis and interpretation, reviewed intellectual content, and approved fnal manuscript. Nuria Rodríguez Núñez performed data analysis and interpretation, reviewed intellectual content, and approved fnal manuscript. Carlos Zamarrón performed data analysis and interpretation, reviewed intellectual content, and approved fnal manuscript. Felipe Calle performed data analysis and interpretation, reviewed intellectual content, and approved fnal manuscript. Francisco Gude carried out analysis and interpretation of the data, economic analysis, reviewed intellectual content, and approved fnal manuscript. Luis Valdés conceived and designed the study, performed data analysis and interpretation, reviewed intellectual content, and approved fnal manuscript.