The duration of antibiotic treatment in community-acquired pneumonia (CAP) lasts about 9-10 days, and is determined empirically. The last North American guideline for CAP recommends using clinical stability criteria as a reference to establish the duration of antibiotic treatment, which would result in about 5 days of antibiotic use for the majority of pneumonia cases. In order to validate this proposal we propose to carry out a randomized multicenter double-blind (until the 5th day) clinical trial with adult CAP patients admitted to 4 hospitals in Euskadi. A control group (with routine treatment) will be compared with an intervention group (antibiotic treatment for at least 5 days, which will be interrupted if temperature is =< 37,8ºC for at least 48 hours and no more than one sign of clinical instability is assessed), with regards to: mortality at 15 days, clinical recovery by days 10 and 30, clinical improvement after days 5 and 10 as evaluated by PRO scales, duration of antibiotic treatment. A non-inferiority dichotomous sequential analysis will be performed (for mortality after 15 days, clinical recovery by day 10 and in follow-up at 30 days, clinical improvement after days 5 and 10, with PRO scales) as well as a superiority analysis for the duration of the antibiotic treatment. A total of 1100 patients will be recruited, following their signed consent, during the inclusion period (18 months). Stability criteria will be measured daily. The rest of the variables will be measured at admission and by telephone on days 10 and 30.

Work hypothesis: We aim to carry out a comparative study between 2 groups of patients (control group, guideline group): in the control group the standard duration of antibiotic treatment will be applied; in the guideline group (intervention group) the duration of antibiotic treatment will be based upon the IDSA/ATS 2007 provisions (ref. 18), which are related to the time needed to reach clinical stability.

1. - Patients who follow the treatment protocol of the so-called control group, with routine treatment, in comparison to the clinical guideline group, will show similar results (sequential dichotomous non-inferiority analysis) in the following clinical parameters: A. Adverse results: 1st, mortality for any cause within the first 15 days after admission; 2nd, development of (major or minor) complications during the first 15 days after admission; B. Beneficial results: 1st, clinical recovery by the 10th day of treatment and in follow-up at 30 days; 2nd, clinical improvement on the 5th and 10th days of treatment, as evaluated by PRO scales.
2. - Patients who follow the treatment protocol of the so-called guideline group (intervention group), in comparison with the control group, with routine treatment, will show better results (superiority analysis) in the following parameters: c. Duration of antibiotic treatment as measured using 2 variables: 1.- the "duration of antibiotic treatment" after admission, expressed in days; 2.- "antibiotic-free days", defined as a period of at least 24 hours in which antibiotics are not given in the 30 day period after admission.

Goals: To validate the recommendations of the latest IDSA/ATS clinical guideline on duration of antibiotic treatment in pneumonia.

Main goals. To compare two treatment routines (control group with routine treatment vs clinical guideline group) based on results in the following clinical parameters:

A. Adverse results: 1st, mortality for any cause within the first 15 days after admission; 2nd, development of (major or minor) complications during the first 15 days after admission; B. Beneficial results: 1st, clinical recovery by the 10th day of treatment and in follow-up at 30 days; 2nd, clinical improvement on the 5th and 10th days of treatment, as evaluated by PRO scales.

2.- Patients who follow the treatment schema of the so-called guideline group (intervention group), in comparison with the control group, with routine treatment, will show better results (superiority analysis) in the following parameters: c. Duration of antibiotic treatment as measured using 2 variables: 1.- the "duration of antibiotic treatment" after admission, expressed in days; 2.- "antibiotic-free days", defined as a period of at least 24 hours in which antibiotics are not given in the 30 day period after admission.

Secondary goals: To compare a control group with routine treatment vs the clinical guideline group based on the following parameters:

1. In-hospital mortality for any cause. Mortality within 30 days after admission, whether related or unrelated to pneumonia.
2. Readmission within 30 days after admission for reasons related or unrelated to pneumonia.
3. Duration of hospital stay.
4. Days to return to normal activity.
5. Recurrence of initial infection or superinfection with need for additional or alternative antibiotics, within 30 days after admission.

Study design - Population under study. Inclusion/exclusion criteria - Main and secondary variables - Data analysis - Ethical and legal aspects (BCP compliance, insurance…) - Aspects related to the research medication (manufacturing, labeling, distribution…) - Funding sources of the medication Randomized multicenter double-blind (until the 5th day) clinical trial involving adult PAC patients (≥18 years of age) admitted to 4 teaching hospitals in Euskadi. All CAP patients hospitalized within a period of 18 months (October-April) will be evaluated for their eligibility. The study involves an intervention and will be undertaken in 2 parallel groups, where the duration of antibiotic treatment following routine clinical practice (control group) will be compared with its duration under the IDSA/ATS provisions for CAP (ref. 18) ("clinical guideline" group). Patients will be randomly assigned to the control group or the "clinical guideline" group on the third day of their hospital stay. Researchers will not be notified of the random assignment until day 5. In the control group patients will be treated according to routine practice for 9 days or longer. In this group there is no interference with the duration of antibiotic treatment. In the "clinical guideline" group patents will receive antibiotics for a minimum of 5 days, antibiotics will be discontinued if temperature is at ≤ 37,8ºC for 48 hours and no more than one sign of clinical instability (see section on definitions) is attested. Once this situation of stability is attained, at each of the participating hospitals a research team doctor will instruct the physician in charge of each patient to drop antibiotic treatment from that moment on.

In patients of the "clinical guideline" group the degree of compliance with the study algorythm will be noted and the reasons that justify skipping it if the case presents itself. The evolution of the clinical trial will be monitored by an independent monitoring team.

Research team members will not be involved in decisions related to discharge from the hospital. After hospital discharge all patient care will depend on the correspoding family physicians. A control visit will be scheduled for the 30th day after admission.

Study location. Multicenter study that will take place in 4 public network Bizkaian hospitals: Galdakao-Usansolo Hospital, Basurto Hospital, Txagorritxu Hospital and San Eloy Hospital.

Study subjects. For a period of 18 months all adult patients, 18 or older, admitted with CAP will be included sequentially. Pneumonia is defined as pulmonary infiltrate shown in a thoracic X-ray not known to be old, plus symptoms consistent with pneumonia, such as cough, dysnea, fever and/or pleuritic chest pain. The following patients will be excluded: patients with human immunodeficiency virus infection, immunosupressed patients (with a solid organ transplant, spelenectomy, treated with a prednisone dose of 10 mg/day or equivalent for longer than 30 days or with other immunodepressors, with neutropenia), those hospitalized in the 14 days prior and those living in assisted care facilities. Pneumonia cases caused by infrequent agents (i.e. P. aerouginosa, S. Aureus) will also be excluded, as well as infectious processes requiring an extended treatment with antibiotocs (i.e. bacterial endocarditis, abscesses), pneumonia cases with pleural effusion requiring a drainage tube, patients who were deceased or admitted to ICU before randomization and those not giving their informed consent.

A total of 1100 cases will be included during the period of study. Losses: those cases not registering the data corresponding to the main goals (loss estimate <5%). A comparative analysis between lost and non-lost patients. For that purpose socio-demographic and clinical variables will be included, as well as those related to severity at admission and the Charlson index (Charlson ME et al. J Chronic Dis 1987;40:373-83).

Ethical aspects. All patients will be informed about the study and they will be asked to give their informed consent (see annex 2). All data will be treated following the law for the protection of personal data (Organic Law 15/1999, 13-12). The study database will be protected with security access measures and the patients' identification data will stay separate from the rest of the information included in the database.

Sample size. We have chosen the equivalence hypothesis in mortality rates between the control group (with conventional guideline) and the intervention group ("clinical guideline group", with treatment template following clinical guideline recommendations). For a unilateral (non-inferiority) contrast (a=0.05, 1- b=0.80) 545 patients will be needed for each group, so as to be able to reject the null hypothesis of non-equivalence of the treatment following the new clinical guidelines as opposed to conventional guidelines. We have assumed an equivalence margin of 1% (http://people.ucalgary.ca/\~patten/blackwelder.html). Sample size has been calculated assuming a 5% rate of losses during follow-up.

Independent variables. 1- severity at admission time. During the 24 hours following arrival at the hospital's Emergency Service the requisite variables will be gathered so as to calculate risk class as established by the Pneumonia Severity Index (PSI) (Fine MJ, et al. N Engl J Med 1997;36:243-50) and the CURB65 scale (Confusion, Urea nitrogen, Respiratory rate, Blood pressure, Age) (Lim WS, et al. Thorax 2003;58:377-82); 2- comorbidities; 3-Charlson index; 4- treatment given: antibiotic prior to admission; adherence of antibiotic treatment to SEPAR provisions; initiation of antibiotic treatment in the first 8 hours; transition from intravenous to oral medication; use of non-invasive mechanical ventilation; 5- bacteriological diagnosis. During the acute phase two hemocultures, a sputum culture and urinary antigen tests will be included, while blood tests for atypical bacteria will be performed during the acute and remission phases.

Dependent variables. 1- mortality for any cause in a 15 day period after admission. The in-hospital mortality and at 30 days will also be analyzed. 2- clinical recovery on the 10th (recovery test - see definitions section) and 30th days; 3- clinical evolution on the 5th and 10th treatment days, measured by PRO scales; a specific validated 18-item questionnaire will be used (Lamping DL, et al. Chest 2002;122:920-9); 4- duration of antibiotic treatment after the time of hospitalization (in days) and days without antibiotic (periods of 24 hours without antibiotic) until the 30th day. 5- hospital readmission within 30 days. 6- duration of hospital stay. 7- clinical recovery (number of days); 8- recurrence of the initial infection or superinfection with need for additional or alternative antibiotic treatment (within 30 days); 9- number of days off work due to disease (within 30 days); 10- number of days with adverse effects from the medication (within 30 days); 11- number of days with the usual (recreation or work) activity restricted by pneumonia (within 30 days); 12- in-hospital complications and at 30 days; 13- need for additional medication.

Information collection. Patients will be included in the study once they are admitted with PAC. A protocol for data collection in both study groups will be designed. Clinical stability criteria will be collected daily twice a day during the admission period. The other study variables will be collected during hospital stay and by means of a structured telephone interview on the 10th (unless the patient is still hospitalized) and 30th days. All the data will be collected by trained blinded staff.

Antibiotic treatment. All patients shall be treated according to SEPAR recommendations (Arch Bronconeumol 2010;46:543-58): group inpatients, not in ICU. Intravenous beta-lactam (amoxicillin/clavulanate or cefotaxime or ceftriasone) plus intravenous macrolide (azitromicyn or claritromicyn), or intravenous levofloxacine; group ICU-admitted patients. Intravenous beta-lactam (cefotaxime or ceftriasone) plus intravenous macrolide (azitromycin or claritromycin), or intravenous beta-lactam (cefotaxime or ceftriasone) plus intravenous levofloxacine. The Research Team will insist on starting antibiotic treatment as soon as possible and on patients being treated with fluoroquinolone or beta-lactam with macrolides. The ultimate decision corresponds to the physician in charge of the patient.

Definitions. Clinical instability: systolic arterial pressure <90 mmHg; breathing rate > 24 breath/min; heart rate >100 heartbeats/min; O2 saturation <90% or O2 arterial pressure <60mmHg. Patients receiving oxygen supplement will be considered stable with O2 ≥95% saturation. Clinical recovery (cure test): resolution or improvement of symptoms and clinical signs related to pneumonia without a need for additional or alternative antibiotic treatment. Clinical recovery on follow-up at 30 days: when there is a continuance of resolution or improvement of symptoms and clinical signs related to pneumonia without a need for additional or alternative antibiotic treatment. Recurrence: new (or worsening) symptoms and signs related to pneumonia and with a new infection of the respiratory tract in a patient considered cured on the 10th day visit.

Statistical analysis: For the descriptive analysis we will use averages, standard deviations, medians and ranks for the quantitave variables and percentages for the qualitative ones. The fundamental analyses will include: 1) first, and in order to make sure that randomization has been done correctly, the fundamental basal variables of the study will be compared between the intervention group and the control group to verify that there are no statistically significant differences between both groups in those variables. Intermediate analyses will be performed every three months of follow-up to see whether there are differences between both groups that may justify discontinuing the study. 2) To respond to the fundamental goals and hypotheses of the study univariate analyses will be performed comparing all the main and secondary dependent variables between the two groups of patients (main independent variable). A subanalysis with the most severe patients will also be conducted (PSI IV and V).

For the comparison of the quantitative variables between the control group and the intervention group Student's t test or Wilcoxon's non-parametric test will be used, depending on whether the studied variable follows a normal distribution or not. For the comparison of the qualitative variables a Chi-square test or Fisher's exact test will be used. In adition a comparative analysis of the basal characteristics of the lost and non-lost patients will be performed, using the same statistical tests previously mentioned.

Problems and Limitations of the Study: 1- As in every clinical trial, the main problem may be our patients' lack of interest to become a part of our study, although we will attempt to reduce the impact of this factor by explaining the study objectives, with its possible benefits and risks and its general interest, both for them and for the rest of pneumonia patients; 2- In the "clinical guideline" group the ultimate decision to discontinue antibiotic treatment after reaching clinical stability criteria will depend on the physician in charge of each patient. It is possible we may find occasional resistance to conform with the established protocol. In order to neutralize protocol transgressions as much as possible, the decision to discontinue antibiotic treatment will be analyzed previously at each hospital by a member of the research team, who will relay it to the physician in charge. This way we will on the one hand ensure the compliance of each case to what is established in the clinical trial, and on the other, encourage the individual physician's cooperation. At any rate, if the protocol is skipped, the reasons surrounding the case will be recorded. 3- Multicenter studies are complicated from an operational point of view. There is the possibility that the ratio between eligible and included patients might differ from one location to the other(s), or that the number of cases included bears no relationship with the size of the service population. In order to neutralize this effect we will introduce the hospital factor in our statistical analyses, besides studying each hospital's caseload separately and comparing them all.

Inclusion Criteria:

• 18 or older,
• admitted with CAP will be included sequentially. Pneumonia is defined as pulmonary infiltrate shown in a thoracic X-ray not known to be old,

Exclusion Criteria:

• patients with human immunodeficiency virus infection,
• immunosupressed patients (with a solid organ transplant, spelenectomy, treated with a prednisone dose of 10 mg/day or equivalent for longer than 30 days or with other immunodepressors, with neutropenia),
• those hospitalized in the 14 days prior and those living in assisted care facilities.
• pneumonia cases caused by infrequent agents (i.e. P. aerouginosa, S. Aureus) will also be excluded, as well as infectious processes requiring an extended treatment with antibiotocs (i.e. bacterial endocarditis, abscesses),
• pneumonia cases with pleural effusion requiring a drainage tube,
• patients who were deceased or admitted to ICU before randomization and those not giving their informed consent.