Quintana et al. BMC Cancer (2016) 16:435 DOI 10.1186/s12885-016-2475-y STUDY PROTOCOL Open Access Colorectal cancer health services research study protocol: the CCR-CARESS observational prospective cohort project José M. Quintana1,8* , Nerea Gonzalez1,8, Ane Anton-Ladislao1,8, Maximino Redondo2,8, Marisa Bare3,8, Nerea Fernandez de Larrea4,8, Eduardo Briones5, Antonio Escobar6,8, Cristina Sarasqueta7,8, Susana Garcia-Gutierrez1,8, Urko Aguirre1,8 and for the REDISSEC-CARESS/CCR group Abstract Background: Colorectal cancers are one of the most common forms of malignancy worldwide. But two significant areas of research less studied deserve attention: health services use and development of patient stratification risk tools for these patients. Methods: Design: a prospective multicenter cohort study with a follow up period of up to 5 years after surgical intervention. Participant centers: 22 hospitals representing six autonomous communities of Spain. Participants/Study population: Patients diagnosed with colorectal cancer that have undergone surgical intervention and have consented to participate in the study between June 2010 and December 2012. Variables collected include pre-intervention background, sociodemographic parameters, hospital admission records, biological and clinical parameters, treatment information, and outcomes up to 5 years after surgical intervention. Patients completed the following questionnaires prior to surgery and in the follow up period: EuroQol-5D, EORTC QLQ-C30 (The European Organization for Research and Treatment of Cancer quality of life questionnaire) and QLQ-CR29 (module for colorectal cancer), the Duke Functional Social Support Questionnaire, the Hospital Anxiety and Depression Scale, and the Barthel Index. The main endpoints of the study are mortality, tumor recurrence, major complications, readmissions, and changes in health-related quality of life at 30 days and at 1, 2, 3 and 5 years after surgical intervention. Statistical analysis: In relation to the different endpoints, predictive models will be used by means of multivariate logistic models, Cox or linear mixed-effects regression models. Simulation models for the prediction of discrete events in the long term will also be used, and an economic evaluation of different treatment strategies will be performed through the use of generalized linear models. Discussion: The identification of potential risk factors for adverse events may help clinicians in the clinical decision making process. Also, the follow up by 5 years of this large cohort of patients may provide useful information to answer different health services research questions. Trial registration: ClinicalTrials.gov Identifier: NCT02488161. Registration date: June 16, 2015. Keywords: Colon cancer, Rectal cancer, Clinical prediction rule, Health services research * Correspondence: josemaria.quintanalopez@osakidetza.eus 1 Unidad de Investigación, Hospital Galdakao-Usansolo, Galdakao, Bizkaia, Spain 8 Red de Investigación en Servicios de Salud en Enfermedades Crónicas (REDISSEC), Galdakao, Bizkaia, Spain Full list of author information is available at the end of the article © 2016 The Author(s). Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Quintana et al. BMC Cancer (2016) 16:435 Page 2 of 12 Background With the purpose of reducing publication bias and im- Colorectal cancers are currently among the most com- proving reproducibility and to serve as reference for fu- mon cancers in both women and men [1, 2]. Although ture manuscripts derived from this study and for readers many scientific approaches have been applied in this we present the study protocol of this multicenter re- field, primarily with respect to diagnosis and treatment, search work. two significant aspects require increased focus, the first of which is the development of clinical prediction rules Methods/design to predict adverse events after surgical treatment of Study design colorectal cancer patients. While some measurements of An observational analytic prospective cohort study with short-term outcomes have been developed, such as the a planned patient follow up period of 5 years following different versions of the Physiological and Operative Se- surgical intervention. verity Score for the enUmeration of Mortality and mor- Setting: patients recruited from 22 hospitals represent- bidity (POSSUM) scoring [3–5], most are not properly ing 9 provinces in Spain, all of which operate under the validated in other settings. Furthermore, prediction Spanish National Health Service (SNHS), which is re- models for medium-term follow up (e.g., 1 or 2 years), a sponsible for the majority of the national population. All period in which the majority of adverse outcomes after applicable residents have free access to their primary treatments are observed, are inadequate. Not only robust care physician and to the hospital ED. Participating hos- clinical outcomes such as mortality, major complica- pitals are: Hospital de Antequera, Hospital Costa del Sol, tions, relapses or reinterventions should be studied and Hospital Universitario de Valme, Hospital Universitario related factors identified, but also the determinants of Virgen del Rocío, Hospital Universitario Virgen de las changes in health-related quality of life, as perceived by Nieves, Hospital Universitario de Canarias, Corporació the patient; these are referred to as patient reported out- Sanitaria Parc Taulí, Althaia, Hospital del Mar, Hospital come measurements (PROMs) [6–8]. The second signifi- Clínico San Carlos, Hospital Universitario La Paz, Hos- cant area of research interest is related to health services pital Infanta Sof ía, Hospital Universitario Fundación research, where multiple factors remain unquantified. Alcorcón, Hospital Galdakao-Usansolo, Hospital Univer- Among these are the quality of the health care process sitario Araba, Hospital Universitario Basurto, Hospital from the time that initial symptoms present in the pa- Universitario Cruces, Hospital Hospitalario Donostia, tient to the time at which the patient receives treatment, Hospital Bidasoa, Hospital de Mendaro, Hospital de equity in access to diagnosis and treatment procedures, Zumarraga and Hospital Universitario Doctor Peset. provision of auxiliary support services (including key Research population: Patients diagnosed with colorec- psychosocial services) or the cost and effectiveness of tal cancer that underwent surgical intervention at one of different surgical approaches [9–11]. the listed hospitals between June 2010 and December Study protocol manuscripts are quite common in the 2012. Patients were informed of the study objectives and cancer research field but mainly to present clinical trials invited to voluntarily participate. Patients were included of treatments for these patients. Less common, though in the study sequentially. no rare, other large and ambitious studies dealing with Expected duration of the study: patients will be quality of care o even health services research have also followed until 5 years after the surgical index interven- published their study protocol. [12, 13]. tion, or until death. Study goals and objectives Selection criteria The purpose of this study is to clarify some of the previ- Patients will be deemed eligible for the study if they have ously stated questions and unknown factors. The specific undergone surgery for colorectal cancer in one of the study objectives are displayed in Fig. 1 and are: 1. To de- participating hospitals. Colorectal cancer diagnosis is termine risk factors for death or major complications, based on anatomopathological diagnosis after a biopsy reoperations, or readmissions in the short-term; 2. To by colonoscopy [10, 14, 15]. determine risk factors for death, tumor recurrence, Inclusion criteria: diagnosis with colon cancer (up to major complications, readmission or deterioration in 15 cm above the anal margin) and rectum (between the quality of life in the mid-term; 3. to evaluate patient re- anal margin and 15 cm above it), where curative or pal- ported outcomes from before intervention to the end of liative surgery by first time was applied. Exclusion cri- the follow-up period. 4. To evaluate the role of bio- teria are colon or rectum in situ cancer, inoperable logical markers in the prediction of adverse results; and tumor, severe mental or physical condition which pre- 5. from a health services research perspective, to evalu- vents the patient from responding to questionnaires, ter- ate equity, appropriateness, access to care, treatment minal illness, inability to respond to questionnaires for cost, and psychosocial support for patients. any reason, or lack of consent to participate in the study. Quintana et al. BMC Cancer (2016) 16:435 Page 3 of 12 Fig. 1 Study objectives Figure 2 illustrates a flow chart of the recruitment Index [16], bowel-specific comorbidities, diagnostic and process. Recruitment data from participating hospitals as screening tests, pre-intervention treatments, and dur- of December 2012 are provided in Table 1. ation between relevant time points in the diagnostic process), preoperative data [including analytical, tumor Methodology markers, diagnostic tests and pre-intervention clinical Eligible patients were primarily identified from the surgi- staging (pre-TNM), and tumor site], data from the out- cal waiting lists of participating hospitals. For those receiv- patient anesthesia visit prior to surgical intervention ing urgent surgery, identification took place during the (ASA stage [17], physiological and cardio-respiratory pa- hospitalization period. Depending on the hospital, one of rameters and tests, and consciousness level based on the two strategies were implemented for the inclusion of Glasgow scale [18]). Variables from the different versions patients: a) surgeons or other healthcare professionals ex- of the POSSUM were specifically included to test all plained the study objectives and invited patients to volun- those rules. tarily participate, during a clinical visit prior to surgery or during hospitalization for surgery; or b) written informa- Data related to hospital admission tion about the goals of the study and the consent form for participation were sent to patients prior to hospitalization a. Data on the surgical intervention (including type of for surgery. intervention, surgical severity, aggravate diagnosis, Following the patient selection process, clinical data type of anesthesia, antibiotic prophylaxis, surgical and PROMs were collected at specific time points. approach, type of surgery, invasion of adjacent Figure 3 summarizes the data-gathering process. organs, additional surgical procedures associated to local extension of the disease, intra surgical Data collection complications, complication grade, duration of A. Medical records and direct data collection surgery, and other information specific to the colon- or recto-surgical intervention), and Intensive Care Clinical data were gathered from medical records and Unit (ICU) period information (duration, mechanical databases by qualified reviewers. A data collection form ventilation requirements, parenteral nutrition, anti- and instruction manual were developed to guide the data biotics, presence of complications, and death). collection process to ensure consistency among centers b. Anatomical pathology data, including tumor-free and reviewers. margin, histologic diagnosis, pre-TNM (pTNM), Upon hospital admission, baseline data were collected lymph nodes analyzed and affected, degree of tumor that included information on patient sociodemographics, differentiation, invasion, remission, and biological clinical data (including data from the initial diagnosis, markers. the pre-intervention process, onset of symptoms, habits and lifestyles (physical activities, alcohol intake or smok- In relation to the biological marker objective of the ing habit), personal and family background, general co- study, 800 paraffin-embedded samples were collected for morbidity data as recorded in the Charlson Comorbidity immunohistochemical analysis to identify clinical and Quintana et al. BMC Cancer (2016) 16:435 Page 4 of 12 Fig. 2 Flow chart of the recruitment process pathological features associated with changes prolifera- Table 1 Recruited patients by area and hospital tion, apoptosis and angiogenesis, and the role of these Autonomous Hospital Valid patients changes in the occurrence of adverse results such as community n cancer relapse and mortality. Samples were collected Andalucía Antequera 41 from patients attending the following hospitals: Hospital Costa del Sol 95 de Antequera, Hospital Costa del Sol, Hospital de Valme, n = 490 Valme 137 Hospital Virgen del Rocío, Hospital Virgen de las Virgen de las Nieves 31 Nieves, Hospital Galdakao-Usansolo, and Hospital Basurto. Virgen del Rocío 186 Canarias n = 101 Complejo Hospitalario de Canarias 101 c. Other data related to hospital admission (length of Cataluña Corporació Parc Taulí 332 stay, presence and degree of complications, n = 689 Althaia 100 treatment information, need for reintervention, or Hospital del Mar 257 death). Complications were grouped in the following Madrid La Paz 169 categories: surgical, medical, infectious, and hematological (hemorrhage/thrombosis/embolism) n = 343 Infanta Sofía 64 (see Additional file 1: Table S1). Clínico de San Carlos 39 Alcorcón 71 Data were subsequently collected up to 30 days after País Vasco Txagorritxu 88 surgery (analytics, diagnostic tests, presence of complica- Bidasoa 33 tions, readmissions, reintervention, social service re- Donostia 245 quirements, or death). Oncologic treatment information is shown in the adittional files section as Table 2. Mendaro 28 Information was next collected through the first year n = 998 Zumarraga 39 of follow-up, including need for and type of treatment Basurto 229 (separately evaluated for colon or rectal tumors), radi- Cruces 139 ation therapy, chemotherapy (treatment schedule, neo or Galdakao-Usansolo 197 adjuvant therapy, number of cycles, presence and degree Valencia n = 128 Doctor Pesset 128 of treatment complications, and supportive care require- ments), other medical treatments and psychological sup- Total 2749 port. Information was also collected on laboratory and Quintana et al. BMC Cancer (2016) 16:435 Page 5 of 12 Fig. 3 Data gathering process from baseline to 5 years of follow up. * Patient reported outcomes measures (PROMs) used: EuroQol-5D (EQ-5D), EORTC QLQ-C30 (European Organization for Research and Treatment of Cancer Quality of Life Questionnaire-C30); QLQ-CR29 (European Organization for Research and Treatment of Cancer Questionnaire Module for Colorectal Cancer), the Duke-UNC Functional Social Support Questionnaire (FSSQ), the Hospital Anxiety and Depression Scale (HADS), and the activities of daily living (ADL) Barthel Index (BI). Patient reported outcome at 30 days were collected in a sub-sample of the overall cohort and after 2 years in a selection of the original hospitals. ICU: Intensive Care Unit diagnostic tests performed, presence of complications, The EQ-5D is a generic health-related quality of life additional surgeries, tumor recurrence, readmission or (HRQoL) questionnaire which consists of two parts, the reintervention, and death. first of which is a description of the state of health in Information collected at the 2-, 3- and 5-year visits five dimensions: mobility, self-care, usual activities, pain/ was similar to that of the 1-year visit. discomfort and anxiety/depression. Each of these dimen- sions is measured by three answer options that define B. Patient reported outcomes measures (PROMs) different levels of severity. The second part is a visual analogue scale in which patients rate their health on a Patients completed the following questionnaires prior scale displayed in the form of a thermometer (20 mm), to surgery and in the follow up after surgery: EuroQol- whose ends are 0 (worst imaginable health state) to 100 5D (EQ-5D), EORTC QLQ-C30 (The European (best state imaginable health) [19]. This questionnaire Organization for Research and Treatment of Cancer has been translated into and validated in Spanish [20]. Quality of Life Questionnaire-C30) and QLQ-CR29 The EORTC QLQ-C30 is a HRQoL questionnaire for (European Organization for Research and Treatment of evaluating the for cancer patients undergoing treatment, Cancer Questionnaire Module for Colorectal Cancer), and has shown good validity and reliability in its Spanish the Duke-UNC Functional Social Support Questionnaire adaptation (Cronbach’s alpha coefficient of 0.7). It consists (FSSQ), the Hospital Anxiety and Depression Scale of 30 items, 28 of which have four possible answers (not (HADS), and the activities of daily living (ADL) Barthel at all, a little, quite a bit, very much) and two items with Index (BI). seven response options (a visual analogue scale where one Quintana et al. BMC Cancer (2016) 16:435 Page 6 of 12 Table 2 Basic description of the whole sample at baseline N (%) Main cancer location Rectal Colon Total 2749 770 (28.01) 1979 (71.99) Gender (Men) 1749 (63.62) 522 (67.79) 1227 (62.00) Agea 68.50 (10.97) 67.14 (11.26) 69.02 (10.81) a Charlson Comorbidity index 2.88 (1.29) 2.83 (1.22) 2.90 (1.32) ASA index I,II 1548 (57.92) 450 (59.92) 1098 (57.13) III 1020 (38.16) 277 (36.88) 743 (38.66) IV 105 (3.93) 24 (3.20) 81 (4.21) Surgical approach (main) Open surgery 1139 (41.89) 307 (40.55) 832 (42.41) Laparoscopy 1580 (58.11) 450 (59.45) 1130 (57.59) Laterality of the tumor (Right –transverse side) 834 (42.14) NA 834 (42.14) pTNM 0, I, II 1581 (57.85) 485 (63.32) 1096 (55.72) III 892 (32.64) 221 (28.85) 671 (34.11) IV 260 (9.51) 60 (7.83) 200 (10.17) Laterality of the tumor: Right-transverse side (appendix, cecum, ascending colon, right hepatic flexure and transverse) versus left side (left splenic flexure, descending colon, and sigmoid colon) ASA, American Society of Anesthesiologists pTNM pathological tumor-node-metastasis staging a Means and, in brackets, standard deviation is “very poor” and seven “excellent”) [21, 22]. The time- The HADS questionnaire is designed specifically for frame assessed refers to the previous week. The scores on individuals with a physical illness. It is divided into two each scale are transformed so that the final values are be- subscales, with seven questions pertaining to symptoms tween 0 and 100. In addition to this overall score, the in- associated with anxiety and seven with depression. Each strument consists of three scales: Global health status, in of the 14 items consists of a 4-point Likert scale (ranging which a high score represents a high performance in from 0 to 3) that applies to the previous week. A total terms of overall health; Functional–scale, which consists score for each subscale is then calculated, ranging from of five subscales of operation: physical, role, emotional, 0 to 21 [25]. The HADS has been translated into and cognitive and social; and a Symptoms scale, describing a validated in Spanish [26]. number of symptoms: fatigue, nausea, vomiting, pain, dys- The Duke-UNC FSSQ questionnaire is composed of 11 pnea, insomnia, decreased appetite, constipation, diarrhea items, each of which is rated on a five-point Likert scale and financial difficulties. A high score on each of these ranging from 1 to 5: the higher the score, the better the scales represents a high level of symptoms or problems in perceived social support. It assesses subjective social sup- the patient. The QLQ-CR29 questionnaire was developed port in two domains, Confidant support: 6 items (score following revision of the QLQ-CR38, and has been dem- range 6–30) and Affective support: 5 items (score range onstrated internationally to have both sufficient validity 5–25), and provides an overall social support measure and reliability to support its use as a supplement to the (score range: 11–55) [27]. The questionnaire has been EORTC QLQ-C30 for assessing patient-reported out- translated into and validated in Spanish [28]. comes during treatment for colorectal cancer in clinical The BI questionnaire consists of 10 items that as- trials and other settings. The QLQ-CR29 questionnaire sess the ability to perform certain basic activities of contains 29 items, 18 of which address gastrointestinal daily life without help. It includes 2- and 4-point re- symptoms, pain and problems with micturition, and sep- sponse options that produce a score ranging from 0 arate scales are available for participants with or without a (completely dependent) to 100 (completely independ- stoma, while separate items address sexual function in ent), with intervals of 5 points [29]. The BI has been men or women. The response categories for each item are translated into and validated in Spanish [30]. the same as those used in the QLQ-C30. This question- Patients were first contacted upon inclusion in the sur- naire has also been validated in Spain [23, 24]. gical waiting list for a surgical intervention or were Quintana et al. BMC Cancer (2016) 16:435 Page 7 of 12 interviewed during their hospital admission period. dependent variable of interest (in this case: mortality, Questionnaires were subsequently provided at 30 days, major complications, relapses or readmissions) for each and at 1, 2, 3 and 5 years after surgery. In all instances, a independent variable included in the multivariate logistic reminder letter was sent to patients who had not replied regression model [31, 32]. Therefore, we estimated that after 15 days. Those who had not responded after a fur- at least 100 events of the dependent variable in the sam- ther 15 days received another reminder letter and, where ple are required in order to ensure that the regression feasible, contacted by telephone. model would adequately converge. Previous data indi- Questionnaires for the 30 day follow up visit were sent to cate that the number of events of the dependent variable patients attending the following hospitals: Hospital Costa mortality would be > 15 % of patients operated on in the del Sol, Hospital Valme, Hospital Virgen del Rocío, Hospital first year, higher that the expected percentages of other Virgen de las Nieves, Corporació Sanitaria Parc Taulí, parameters. We therefore estimated that more than 300 Hospital Universitario La Paz, Hospital Infanta Sofía, events of any of the dependent variables of interest Hospital Clínico San Carlos, Hospital Fundación Alcorcón, should be included. Hospital Basurto, Hospital Cruces, Hospital Donostia, Sampling: we consecutively collected all new cases Hospital Galdakao-Usansolo, Hospital Txagorritxu. until the sample size was achieved. Figure 3 illustrates the data gathering process estab- lished for the entire follow-up period. B. Missing data assumptions and recoding of variables Safety considerations Tumor location There is no explicit intervention in this study. Neverthe- Tumor location was categorized as follows: the appen- less, due to the sensitive nature of the disease and for rea- dix, caecum, ascending colon, hepatic flexure and trans- sons of confidentiality, all information provided to verse colon were defined as the right colon, whereas the patients, either directly by the researchers or reviewers or splenic flexure, descending colon, sigmoid and rectosig- included in the questionnaires, avoid references to cancer moid were defined as the left colon. Rectal tumors were or tumors, referring instead, to “your bowel problem”. Pa- also recorded. tients may refuse to participate at any time during the study, whether during recruitment or follow-up. Further- Symptomatology more, patients who prefer to complete the questionnaire Patient pre-intervention disease-specific symptoms were by phone are permitted to do so. All participating centers collected. Where missing observations were observed for have referring staff available for questions that the patient any recorded symptom, it was assumed to be asymptom- or patient’s family may have about the research project. atic. A new variable was created according to the pres- ence/absence of these symptoms to determine whether Follow-up the patient had experienced any type of symptomatology Follow-up visits up to two years will be performed at all or not. 22 participating hospitals. Due to financial restrictions and organizational resources, 3- and 5-year follow-up visits will be restricted to 18 hospitals (Hospital de Antequera, Complications Hospital Costa del Sol, Hospital Universitario de Valme, A list of several complications (yes/no answers) were Hospital Universitario Virgen del Rocío, Hospital collected throughout the course of the follow-up period Universitario Virgen de las Nieves, Hospital Universitario (intra-surgical, during the hospital admission post- de Canarias, Corporació Sanitaria Parc Taulí, Hospital del surgical, and at 30 days and 1, 2, 3 and 5 years after the Mar, Hospital Clínico San Carlos, Hospital Universitario intervention). If missing data were observed for each La Paz, Hospital Infanta Sofía, Hospital Universitario item in the aforementioned list, no complication was Fundación Alcorcón, Hospital Galdakao-Usansolo, considered to have occurred. Subsequently, at each Hospital Universitario Araba, Hospital Universitario measurement point, an overall dichotomous variable was Basurto, Hospital Universitario Cruces, Hospital defined to classify patients as with and without compli- Hospitalario Donostia, and Hospital Universitario cations. Complications were also classified into different Doctor Peset). types, infectious, hematological, surgical or medical. Data management and statistical analysis Surgical severity A. Sample size selection We redefined the surgical severity variable into three categories (moderate, major or complex major) depend- Studies of predictive model development indicate that ing on the type of procedure used to excise the tumor or is necessary to include at least 10 events of the to palliate symptoms. Quintana et al. BMC Cancer (2016) 16:435 Page 8 of 12 The moderate severity category referred to transanal final performance models. An AUC value > 0.80 will indi- endoscopic microsurgery (TEM) or discharge colostomy; cate that the model discriminates well. Furthermore, the major severity category to surgical techniques such multivariate Cox regression models will be used to answer as right/left hemicolectomy, any colectomy, total procto- to those objectives. On the other hand, in cases of con- colectomy, sigmoidectomy or transanal resection; and tinuous outcomes (variables related to deterioration of the complex major severity category corresponded to quality of life) as with objective 3, the use of multivariate anterior resections or abdominoperineal amputation. hierarchical linear mixed models is proposed. The intra- Any intervention that involved a surgical colorectal tech- class correlation coefficient (ICC) will be calculated to as- nique, in addition to another procedure that affected any sess the correlation among observations within a cluster other organ, was considered a complex major surgical by dividing the patient variance by the total sample vari- intervention. ance. A small coefficient would indicate that patients and We did not consider any colorectal cancer surgeries as participant centers are independent. belonging to the minor severity category. For study endpoints related to health services re- search, the following statistical approaches will be used: pTNM staging for two-level outcomes, the Student’s t-distribution or TNM classification was assessed according to the the non-parametric Wilcoxon test (for non-normal dis- American Joint Committee on Cancer (AJCC), 7th tributions) will be applied. Where there are three or Edition 2010 [14], and focused on the pre-intervention/ more categories in the outcome, ANOVA analysis or a clinical (cTNM) and the histopathological report Kruskal–Wallis approach (for non-normal distribution) (pTNM). For the final stage grouping, pTNM was con- are the proposed tests. Otherwise, for categorical vari- sidered. When only partial pTNM was available, a com- ables, the Chi-square test (or Fisher’s Exact method, bination of cTNM and pTNM data were considered: (a) where required) will be used. Multivariate models will where the pM component was not recorded, pM was re- be used as needed for appropriate adjustments. For placed by the corresponding cM stage or, when the latter some analyses, subgroups based on tumor location was missing, pM was assumed to be pM0; (b) in subjects (colon or rectum) will be identified. whose pN component was unobserved, the pTNM was An analysis of comparative costs between two alterna- classified as at least stage III if there was evidence that tive surgical approaches (open and laparoscopic surgery) lymph nodes were infiltrated. Otherwise, pTNM was will be performed. For this purpose, generalized linear specified as stage II; and (c) tumors without pT value de- models (GLM) will be employed using the patient level scribed pN0-pNx and pM0-pMx parameter values, cost as the dependent variable and taking clinical covari- where pTNM was therefore deemed to be stage II. ates into account (age, sex, stage, among others). The When pTNM was fully unobserved, it was replaced by statistical model will determine which covariates are as- the analogous cTNM data. If the latter information was sociated with higher costs, as well as any significant dif- also missing, it was recorded as a missing value. ferences in cost between the two alternatives. Analysis will be conducted at 2 and 5 years of follow up. C. Statistical Analysis Finally, results related to treatment will be a valuable source of information for inclusion in a discrete-event In relation to the different study objectives, explora- simulation model reproducing a colorectal cancer- tory analysis of the population sample will be performed screening program. Clinical and quality of life data for using mean and standard deviations for continuous vari- treatment of colorectal cancer detected by screening or ables (or median and interquartilic ranges, when the clinical approaches will be useful to assess not only the observed variables do not follow the normality distribu- impact of screening, but also to compare different treat- tion) and frequencies and percentages for qualitative ment strategies through a cost-effectiveness or budget variables. impact analysis. Predictive models will be developed to address objec- tives 1 and 2. The overall sample will be randomly divided Quality assurance into derivation and validation sets. On the one hand, for Reviewers were instructed in the identification of pertin- binary outcomes (such as death, tumor recurrence, major ent data and a specific manual was developed to aid in complications, or readmission), multivariate hierarchical data collection. All data is maintained in an “ad hoc” logistic models will be used. Goodness-of-fit of models database. Personal patient data that may allow identifica- will be assessed by means of the Hosmer–Lemeshow test tion of patients will be maintained separately from the (p-values > 0.05 indicate good model calibration) and the main database, where all clinical and PRO data are in- area under the Receiver Operating Curve (ROC), AUC, cluded. To access the database, specific usernames and will be used to evaluate the discrimination ability of the passwords are required. A log has been developed to Quintana et al. BMC Cancer (2016) 16:435 Page 9 of 12 monitor database access. Databases are located in six sample and one for the subsampling cohort, where bio- places in the servers of participating hospitals. Each hos- logical data was requested. All information was kept pital has a designated person in charge of the supervi- confidential. The Institutional Review Boards of the par- sion of the data collection who serves as a referral ticipating hospitals approved this study. contact for the reviewers. A project manager coordi- nated the study in the initial 2 years, with study coordin- Discussion ation checks performed every 6 months to evaluate the As illustrated in Fig. 2, by the time of manuscript prep- quality and completeness of the data and request add- aration, 2749 patients had been recruited according to itional information and corrections if needed. Twice the study protocol described previously, 1980 of which yearly, the study investigators meet to review the status had been diagnosed with colon cancer and 769 with rec- of data collection and any issues that arise. tal cancer. This population represents a relatively large cohort of colorectal cancer patients for prospective fol- Duration of the project low up. Description of the basic characteristics of the The entire project is planned to last for more than sample, stratified by colon or rectal cancer diagnosis, is 7 years (2 years for recruitment and 5 for follow-up until displayed in Table 2. the end of 2017), until the data from the last patient has been collected and entered into the database. Following this, a 6-month period of error correction and final data- Anticipated issues base preparation is anticipated. Manuscript writing may The main limitations of this type of study design are commence at intervals of 1, 2, 3 and 5 years of follow- those related to missing patients as a result of an exten- up, once errors have been corrected and agreed on by sive follow-up period (5 years, in this instance) and, at the coordination team. the same time, missing data related to some variables, especially those related to patient questionnaires. To Project management minimize these risks, all patients are informed of the ob- Seven study leaders are responsible for each goal and in jectives of the study prior to enrollment and again dur- different geographical areas, along with one coordinator. ing the follow-up visits. Furthermore, to reduce missing Dr M. Bare was tasked with the development of predic- data from questionnaires, up to three letters are sent to tion models to determine risk factors of adverse events patients and conversion to phone visits is available upon in the short-term. Dr J.M. Quintana is responsible for request. Given the lengthy follow up period, regular pa- objectives 2 (to determine risk factors of death, tumor tient contact are established. Furthermore, the contact recurrence, major complications, readmission or deteri- physicians at each hospital remind patients of their par- oration of quality of life in the mid-term) and 3 (evalu- ticipation in the study where possible. ation of patient reported outcomes from before the An additional issue is the reliability of data gathering intervention to the end of the follow-up period). He is by reviewers. To standardize the collection of data and also the coordinator of the entire study. Dr M. Redondo avoid errors, all reviewers are adequately trained and is responsible for objective 4 (to study the role of bio- have received specific instructions on how to collect data logical markers in the prediction of similar adverse re- for each variable. sults). The goals related to health services research are assigned to the following: C. Sarasqueta (equity), E. Expected outcomes of the study Briones (quality of care and appropriateness), A. Escobar The identification of potential risk factors for adverse (evaluation of psychosocial support interventions) and events (death, major complications, relapse, reoperations N. Fernandez-de-Larrea (cost of various surgical ap- or readmission) in the short- and medium-term has un- proaches). Dr M. Redondo and Urko Aguirre, acting as deniable practical value from a clinical perspective. coordinator, are responsible for the follow-up period to These data will assist clinicians to identify and predict 3 and 5 years. The leaders of each study goal, in con- possible unwanted outcomes, establish preventive mea- junction with the coordinator, comprise the coordination sures, if feasible, to correct these outcomes should they committee responsible for all decisions. occur, and provide patients with a medium-term prog- nosis. In addition, the development of simple clinical Ethics prediction rules that are easy-to-implement in daily All patients were informed of the objectives of the study practice on desktop computers, tablet, mobile phones, and invited to voluntarily participate at the first contact and other devices may facilitate clinicians in making de- and at the 3-year follow-up visit. Patients who agreed to cisions to prevent such occurrences. participate provided written consent. Two types of in- The large cohort of patients who have undergone sur- formed consent forms were created, one for the entire gery for colon or rectal cancer followed by 5 years of Quintana et al. BMC Cancer (2016) 16:435 Page 10 of 12 post-surgical follow up may provide useful information 1 Unidad de Investigación. Hospital Galdakao-Usansolo, Galdakao-Bizkaia/Red de to help address other health services research questions. Investigación en Servicios de Salud en Enfermedades Crónicas – REDISSEC. 2 Epidemiologia Clínica y Cribado de Cancer, Corporació Sanitaria Parc Taulí, Sabadell/Red de Investigación en Servicios de Salud en Enfermedades Dissemination of results and publication policy Crónicas – REDISSEC. 3 Servicio de Laboratorio. Hospital Costa del Sol, Málaga/REDISSEC. For the purpose of this study, a research group called 4 Unidad de Epidemiología. Distrito Sevilla, Servicio Andaluz de Salud. REDISSEC CARESS-CCR (Results and Health Services 5 Centro Nacional de Epidemiología, Instituto de Salud Carlos III, Madrid/ Research in Colorectal Cancer)- has been established, REDISSEC. 6 Unidad de Investigación. Hospital Universitario Donostia/Instituto de and reflects the primary research participants of each Investigación Sanitaria Biodonostia, Donostia – REDISSEC. center. For publication policy purposes, an author has to 7 Unidad de Investigación. Hospital Universitario Basurto, Bilbao/REDISSEC. fulfill the necessary requirements by contributing to 8 Servicio de Epidemiología. Hospital Costa del Sol, Málaga – REDISSEC. 9 Department of Preventive Medicine and Public Health, Univesity of Valencia/ each of three activities: 1) conception/design and/or ana- CIBER de Epidemiología y Salud Pública (CIBERESP)/CSISP-FISABIO, Valencia. lysis/interpretation, 2) writing, and 3) approval of final 10 Unidad de Evaluación de Tecnologías Sanitarias, Agencia Laín Entralgo, Madrid. 11 version, and take public responsibility for the content of Unidad Apoyo a Docencia-Investigación. Dirección Técnica Docencia e Investigación. Gerencia Adjunta Planificación. Gerencia de Atención Primaria the paper. All co-authors have to review and agree with de la Consejería de Sanidad de la Comunidad Autónoma de Madrid. the contents of the manuscript as submitted. Due to the 12 Departamento de Matemática Aplicada, Estadística e Investigación study objectives, the entire study and manuscripts will Operativa, UPV- REDISSEC. 13 Epidemiologia Clínica y Cribado de Cancer, Corporació Sanitaria Parc Taulí, follow the STROBE guidelines for conducting and dis- Sabadell/REDISSEC. seminate observational studies and the TRIPOD state- 14 Servicio de Oncología. Hospital Universitario Basurto, Bilbao. 15 ment for reporting of a multivariable prediction model Servicio de Cirugía General. Hospital Universitario Basurto, Bilbao. 16 Servicio de Anatomía Patológica. Hospital Galdakao-Usansolo, Galdakao. for individual prognosis or diagnosis [33, 34]. 17 Servicio de Cirugía General. Hospital Universitario Araba, Vitoria-Gasteiz. The main study results will be disseminated in the 18 Servicio de Gastroenterología. Hospital Universitario de Canarias, La Laguna. 19 media, to clinicians, managers and policy makers in the Servicio de Cirugía General. Hospital Galdakao-Usansolo, Galdakao. 20 IMAS-Hospital del Mar, Barcelona. proper format. 21 Servicio de Cirugía General y Digestiva. Hospital Universitario Donostia. 22 Instituto de Investigación Sanitaria Biodonostia, Donostia. 23 Additional file Servicio de Cirugía General y Digestiva. Hospital de Mendaro. 24 Servicio de Cirugía General y Digestiva. Hospital de Zumárraga. 25 Servicio de Cirugía General y Digestiva. Hospital del Bidasoa. Additional file 1: Table S1. Grouping of complications. Table S2. 26 Servicio de Oncología Médica. Hospital Universitario Donostia. Chemotherapy information at the beginning of the study. (DOCX 17 kb) 27 Instituto de Biomedicina de Sevilla. Hospital Universitario Virgen del Rocío, Sevilla. 28 Servicio de Cirugía. Hospital Universitario Virgen de Valme, Sevilla. Acknowledgments 29 Servicio de Cirugía General y Aparato Digestivo. Hospital Dr. Pesset, Valencia. We are grateful to the participant patients who voluntarily took part in this 30 Servicio de Cirugía General y Aparato Digestivo. Agencia Sanitaria Costa del study. We also gratefully thank the doctors and all the interviewers from the Sol, Marbella. participating hospitals (Hospital de Antequera, Hospital Costa del Sol, 31 Servicio de Oncología Médica. Agencia Sanitaria Costa del Sol, Marbella. Hospital Universitario de Valme, Hospital Universitario Virgen del Rocío, 32 Servicio de Cirugía. Hospital de Antequera. Hospital Universitario Virgen de las Nieves, Hospital Universitario de Canarias, 33 Coloproctology Unit, General and Digestive Surgery Service, Corporació Corporació Sanitaria Parc Taulí, Althaia, Hospital del Mar, Hospital Clínico San Sanitaria Parc Taulí, Sabadell. Carlos, Hospital Universitario La Paz, Hospital Infanta Sofía, Hospital 34 Digestive Diseases Department, Corporació Sanitaria Parc Taulí, Sabadell. Universitario Fundación Alcorcón, Hospital Galdakao-Usansolo, Hospital 35 Pathology Service, Corporació Sanitaria Parc Taulí, Sabadell. Universitario Araba, Hospital Universitario Basurto, Hospital Universitario 36 Medical Oncology Department, Corporació Sanitaria Parc Taulí, Sabadell/ Cruces, Hospital Hospitalario Donostia, Hospital Bidasoa, Hospital de Mendaro, REDISSEC. Hospital de Zumarraga and Hospital Universitario Doctor Peset), for their 37 General and Digestive Surgery Service, Parc de Salut Mar, Barcelona. invaluable collaboration in patient recruitment, and to the Research 38 General and Digestive Surgery Service, Althaia- Xarxa Assistencial Committee of the participating hospitals. Universitaria, Manresa. The REDISSEC CARESS-CCR (Results and Health Services Research in Colorectal 39 Catalonian Cancer Strategy Unit, Department of Health, Institut Català Cancer)- group d’Oncología. Jose María Quintana1, Marisa Baré2, Maximino Redondo3, Eduardo Briones4, 40 Nerea Fernández de Larrea5, Cristina Sarasqueta6, Antonio Escobar7, Francisco Medical Oncology Department, Institut Català d’Oncología. Rivas8, Maria M. Morales-Suárez-9, Juan Antonio Blasco10, Isabel del Cura11, 41 Agency for Health Quality and Assessment of Catalonia –AquAS- and REDISSEC. Inmaculada Arostegui12, Amaia Bilbao7, Nerea González1, Susana 42 Agency for Health Quality and Assessment of Catalonia –AQuAS-, CIBER de García-Gutiérrez1, Iratxe Lafuente1, Urko Aguirre1, Miren Orive1, Josune Epidemiología y Salud Pública –CIBERESP. 43 Martin1, Ane Antón-Ladislao1, Núria Torà13, Marina Pont13, María Purificación Servicio de Cirugía General y del Aparato Digestivo, Hospital Universitario Martínez del Prado14, Alberto Loizate15, Ignacio Zabalza16, José Errasti17, Fundación Alcorcón, Madrid. 44 Antonio Z Gimeno18, Santiago Lázaro19, Mercè Comas20, Jose María Servicio de Cirugía General y Aparato Digestivo, Hospital Universitario Enríquez21, Carlos Placer21, Amaia Perales22, Iñaki Urkidi23, Jose María Erro24, Clínico San Carlos, Madrid. 45 Enrique Cormenzana25, Adelaida Lacasta26, Pep Piera Pibernat26, Elena Servicio Cirugía General y del Aparato Digestivo, Hospital Universitario Campano27, Ana Isabel Sotelo28, Segundo Gómez-Abril29, F. Medina-Cano30, Infanta Sofía, San Sebastián de los Reyes, Madrid. 46 Julia Alcaide31, Arturo Del Rey-Moreno32, Manuel Jesús Alcántara33, Rafael Servicio de Cirugía General y del Aparato Digestivo, Hospital Universitario Campo34, Alex Casalots35, Carles Pericay36, Maria José Gil37, Miquel Pera37, Pablo La Paz, Madrid. 47 Collera38, Josep Alfons Espinàs39, Mercedes Martínez40, Mireia Espallargues41, REDISSEC. Unidad de Apoyo a la Investigación, Gerencia Asistencial de Caridad Almazán42, Paula Dujovne Lindenbaum43, José María Fernández-Cebrián43, Atención Primaria de la Comunidad de Madrid, Madrid. 48 Rocío Anula Fernández44, Julio Ángel Mayol44, Ramón Cantero45, Héctor REDISSEC. Unidad de Gestión Sanitaria. Hospital del Alto Deba. Guadalajara46, María Heras46, Damián García46, Mariel Morey47, Javier Mar48. Mondragon-Arrasate. Spain. Quintana et al. BMC Cancer (2016) 16:435 Page 11 of 12 Funding Author details 1 This work was supported in part by grants from the Fondo de Investigación Unidad de Investigación, Hospital Galdakao-Usansolo, Galdakao, Bizkaia, Sanitaria (PS09/00314, PS09/00910, PS09/00746, PS09/00805, PI09/90460, Spain. 2Unidad de Investigación, Hospital Costa del Sol, Málaga, Spain. 3 PI09/90490, PI09/90397, PI09/90453, PI09/90441); Department of Health of Unidad de Epidemiología Clínica, Corporacio Parc Tauli, Barcelona, Spain. the Basque Country (2010111098); KRONIKGUNE –Centro de Investigación en 4 Departamento de Salud, Madrid, Spain. 5UDG Salud Publica, Distrito AP Cronicidad (KRONIK 11/006); and the European Regional Development Fund Sevilla, Sevilla, Spain. 6Unidad de Investigación, Hospital Basurto, Bilbao, (ERDF). These institutions had no further role in study design; in the Bizkaia, Spain. 7Unidad de Investigación, Hospital Donosti, Donostia-San collection, analysis and interpretation of data; in the writing of the Sebastian, Gipuzkoa, Spain. 8Red de Investigación en Servicios de Salud en manuscript; nor in the decision to submit the paper for publication. Enfermedades Crónicas (REDISSEC), Galdakao, Bizkaia, Spain. Availability of data and materials Received: 5 February 2016 Accepted: 30 June 2016 It is expected to be available when the study is finished and data is ready to be analyzed. References Authors’ contributions 1. Center MM, Jemal A, Smith RA, Ward E. Worldwide variations in colorectal JMQ participated in the conception and design, obtaining of funding, coordination, analysis and interpretation of the data, drafting of the article, cancer. CA Cancer J Clin. 2009;59:366–78. critical revision of the article for important intellectual content, and final 2. Gellad ZF, Provenzale D. Colorectal cancer: national and international approval of the article. NG participated in the design, provision of patients, perspective on the burden of disease and public health impact. collection and assembly of data, analysis and interpretation of the data, Gastroenterology. 2010;138:2177–90. drafting of the article, critical revision of the article for important intellectual 3. Copeland GP, Jones D, Walters M. POSSUM: a scoring system for surgical content, and final approval of the article. AAL participated in the analysis and audit. Br J Surg. 1991;78:355–60. interpretation of the data, statistical expertise, collection and assembly of 4. Richards CH, Leitch FE, Horgan PG, McMillan DC. A systematic review of data, drafting of the article, critical revision of the article for important POSSUM and its related models as predictors of post-operative mortality intellectual content, and final approval of the article. MR participated in the and morbidity in patients undergoing surgery for colorectal cancer. design, obtaining of funding, provision of patients, collection and assembly J Gastrointest Surg. 2010;14:1511–20. of data, analysis and interpretation of the data, drafting of the article, critical 5. Moonesinghe SR, Mythen MG, Das P, Rowan KM, Grocott MP. Risk revision of the article for important intellectual content, and final approval of stratification tools for predicting morbidity and mortality in adult patients the article. MB participated in the design, obtaining of funding, provision of undergoing major surgery: qualitative systematic review. Anesthesiology. patients, collection and assembly of data, analysis and interpretation of the 2013;119:959–81. data, drafting of the article, critical revision of the article for important 6. Montazeri A. Quality of life data as prognostic indicators of survival in intellectual content, and final approval of the article. NFdL participated in the cancer patients: an overview of the literature from 1982 to 2008. Health design, provision of patients, collection and assembly of data, analysis and Qual Life Outcomes. 2009;7:102. interpretation of the data, drafting of the article, critical revision of the article 7. McNair AG, Whistance RN, Forsythe RO, Rees J, Jones JE, Pullyblank AM, for important intellectual content, and final approval of the article. EB et al. Synthesis and summary of patient-reported outcome measures to participated in the design, obtaining of funding, provision of patients, inform the development of a core outcome set in colorectal cancer surgery. collection and assembly of data, analysis and interpretation of the data, Colorectal Dis. 2015;17:O217–29. drafting of the article, critical revision of the article for important intellectual 8. Jorgensen ML, Young JM, Solomon MJ. Optimal delivery of colorectal content, and final approval of the article. AE participated in the design, cancer follow-up care: improving patient outcomes. Patient Relat Outcome obtaining of funding, provision of patients, collection and assembly of data, Meas. 2015;6:127–38. analysis and interpretation of the data, drafting of the article, critical revision 9. Hill LB, O’Connell JB, Ko CY. Colorectal cancer: epidemiology and health of the article for important intellectual content, and final approval of the services research. Surg Oncol Clin N Am. 2006;15:21–37. article. CS participated in the design, obtaining of funding, provision of 10. Freund KM, Battaglia TA, Calhoun E, Dudley DJ, Fiscella K, Paskett E, et al. patients, collection and assembly of data, analysis and interpretation of the National Cancer Institute Patient Navigation Research Program: methods, data, drafting of the article, critical revision of the article for important protocol, and measures. Cancer. 2008;113:3391–9. intellectual content, and final approval of the article. SGG participated in the 11. Ara R, Basarir H, Keetharuth AD, Barbieri M, Weatherly HL, Sculpher MJ, et al. conception and design, analysis and interpretation of the data, drafting of Are policy decisions on surgical procedures informed by robust economic the article, critical revision of the article for important intellectual content evidence? A systematic review. Int J Technol Assess Health Care. and final approval of the article. UA participated in the conception and 2014;30:381–93. design, obtaining of funding, collection and assembly of data, analysis and 12. Fenlon D, Richardson A, Addington-Hall J, Smith P, Corner J, Winter J, et al. interpretation of the data, statistical expertise, drafting of the article, critical A cohort study of the recovery of health and wellbeing following colorectal revision of the article for important intellectual content and final approval of cancer (CREW study): protocol paper. BMC Health Serv Res. 2012;12:90. the article. All the previous authors have given final approval of the version 13. Esteva M, Ramos M, Cabeza E, Llobera J, Ruiz A, Pita S, et al. Factors of the article to be published. influencing delay in the diagnosis of colorectal cancer: a study protocol. BMC Cancer. 2007;7:86. 14. Edge S, Byrd DR, Compton CC, Fritz AG, Greene FL, Trotti A. American Joint Competing interests Committee on Cancer: AJCC cancer staging manual. 7th ed. New York: The authors declare that they have no competing interests. Springer; 2016. 15. Lopez-Abente G, Ardanaz E, Torrella-Ramos A, Mateos A, Delgado-Sanz C, Consent for publication Chirlaque MD. Changes in colorectal cancer incidence and mortality trends Not applicable. in Spain. Ann Oncol. 2010;21 Suppl 3:iii76–82. 16. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying Ethics approval and consent to participate prognostic comorbidity in longitudinal studies: development and validation. This project was approved by the following Ethics Committees in Spain J Chronic Dis. 1987;40:373–83. (reference number of approval, when provided, in brackets): Specifically, the 17. Hightower CE, Riedel BJ, Feig BW, Morris GS, Ensor Jr JE, Woodruff VD, et al. Ethics Committees of Hospital de Txagorritxu (2009–20), Galdakao, Donostia A pilot study evaluating predictors of postoperative outcomes after major (5/09), Basurto, Marbella (10/09), Ethics Committee of the Basque Country abdominal surgery: physiological capacity compared with the ASA physical (PI2014084). status classification system. Br J Anaesth. 2010;104:465–71. All patients were informed of the objectives of the study and invited to 18. Teasdale G, Jennett B. Assessment of coma and impaired consciousness. A voluntarily participate at the first contact and at the 3-year follow-up visit. practical scale. Lancet. 1974;2:81–4. Patients who agreed to participate provided written consent. 19. Brooks R. EuroQol: the current state of play. Health Policy. 1996;37:53–72. Quintana et al. BMC Cancer (2016) 16:435 Page 12 of 12 20. Badia X, Roset M, Montserrat S, Herdman M, Segura A. [The Spanish version of EuroQol: a description and its applications. European Quality of Life scale]. Med Clin (Barc). 1999;112 Suppl 1:79–85. 21. Aaronson NK, Ahmedzai S, Bergman B, Bullinger M, Cull A, Duez NJ, et al. The European Organization for Research and Treatment of Cancer QLQ-C30: a quality-of-life instrument for use in international clinical trials in oncology. J Natl Cancer Inst. 1993;85:365–76. 22. Arraras JI, Arias F, Tejedor M, Pruja E, Marcos M, Martinez E, et al. The EORTC QLQ-C30 (version 3.0) quality of life questionnaire: validation study for Spain with head and neck cancer patients. Psychooncology. 2002;11:249–56. 23. Whistance RN, Conroy T, Chie W, Costantini A, Sezer O, Koller M, et al. Clinical and psychometric validation of the EORTC QLQ-CR29 questionnaire module to assess health-related quality of life in patients with colorectal cancer. Eur J Cancer. 2009;45:3017–26. 24. Arraras JI, Suarez J, Arias d, V, Vera R, Asin G, Arrazubi V, et al. The EORTC quality of life questionnaire for patients with colorectal cancer: EORTC QLQ- CR29 validation study for spanish patients. Clin Transl Oncol. 2011;13:50–6. 25. Zigmond AS, Snaith RP. The hospital anxiety and depression scale. Acta Psychiatr Scand. 1983;67:361–70. 26. Quintana JM, Padierna A, Esteban C, Arostegui I, Bilbao A, Ruiz I. Evaluation of the psychometric characteristics of the Spanish version of the hospital anxiety and depression scale. Acta Psychiatr Scand. 2003;107:216–21. 27. Broadhead WE, Gehlbach SH, de Gruy FV, Kaplan BH. The Duke-UNC Functional Social Support Questionnaire. Measurement of social support in family medicine patients. Med Care. 1988;26:709–23. 28. Bellon Saameno JA, Delgado SA, Luna del Castillo JD, Lardelli CP. [Validity and reliability of the Duke-UNC-11 questionnaire of functional social support]. Aten Primaria. 1996;18:153–63. 29. MAHONEY FI, BARTHEL DW. Functional evaluation: the Barthel index. Md State Med J. 1965;14:61–5. 30. Cabanero-Martinez MJ, Cabrero-Garcia J, Richart-Martinez M, Munoz- Mendoza CL. The Spanish versions of the Barthel index (BI) and the Katz index (KI) of activities of daily living (ADL): a structured review. Arch Gerontol Geriatr. 2009;49:e77–84. 31. Laupacis A, Sekar N, Stiell IG. Clinical prediction rules. A review and suggested modifications of methodological standards. JAMA. 1997;277:488–94. 32. Vittinghoff E, McCulloch CE. Relaxing the rule of ten events per variable in logistic and Cox regression. Am J Epidemiol. 2007;165:710–8. 33. Von EE, Altman DG, Egger M, Pocock SJ, Gotzsche PC, Vandenbroucke JP. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Ann Intern Med. 2007;147:573–7. 34. Moons KG, Altman DG, Reitsma JB, Ioannidis JP, Macaskill P, Steyerberg EW, et al. Transparent Reporting of a multivariable prediction model for Individual Prognosis or Diagnosis (TRIPOD): explanation and elaboration. Ann Intern Med. 2015;162:W1–73. Submit your next manuscript to BioMed Central and we will help you at every step: • We accept pre-submission inquiries • Our selector tool helps you to find the most relevant journal • We provide round the clock customer support • Convenient online submission • Thorough peer review • Inclusion in PubMed and all major indexing services • Maximum visibility for your research Submit your manuscript at www.biomedcentral.com/submit