On the data to know the prioritization and vulnerability of patients on surgical waiting lists

The data presented in this article are complementary material to our work entitled “A decision support system for prioritization of patients on surgical waiting lists: A biopsychosocial approach”. We prepared, together with physicians, a survey was used in the otorhinolaryngology unit of the Hospital of Talca for a period of five months, between February 05, 2018 and June 29, 2018. Two hundred and five surveys were collected through 20 biopsychosocial criteria, which allowed measuring the priority and vulnerability of patients on the surgical waiting list. The data allow choosing and preparing patients for surgery according to both a dynamic score and a vulnerability level.


Data description
The data set in this article describes biopsychosocial parameters of the patients of the otolaryngology unit on the surgical waiting list. Fig. 1 describes in detail the survey applied to each of the patients who consulted for their disease in the physician's office. With this instrument, nurses and physicians collected the data that would then be analysed to prioritize their patients. Fig. 2 describes the normalization of the parameters, which were transformed before data analysis. Table 1 describes the opinion of physicians concerning patient ages. Table 2 shows central tendency measures of the 20 parameters. Table 3 describes the parameters and how they were incorporated into the data set. Table 4 describes the opinion given to each parameter by each of the physicians. Tables 5e7 describe the   Specifications Table   Subject Health. Specific subject area Priority and vulnerability of patients. Type of data Tables, figures. How data were acquired During the patient's interview, physicians fill a survey that indicates different aspects regarding the condition of the patient. This data is later used for estimating the priority and vulnerability of the surgical waiting list. Data format Raw and analysed. Parameters for data collection The construction of the dataset was done for analysis. The specialists in otolaryngology suggested the process of data collection in the application of their day-to-day activities with patients. Description of data collection The data of surveys were collected in the polyclinic of the otorhinolaryngology unit of the Hospital of Talca for a period of five months, between February 05, 2018 and June 29, 2018. The process of starting the patient's survey was carried out at first by a nurse and then finished by a physician. Then, to prioritize and measure the vulnerability of patients on the surgical waiting list of the otorhinolaryngology unit, we propose a method of data analysis. Specifically, we transform physicians' knowledge into quantitative information that allowed us to; (1) characterize patients and (2)

Value of the Data
The data contain records associated to 20 biopsychosocial parameters from 205 patients. These data (i) include the opinion of seven physicians, (ii) allow the characterization of the patients, (iii) allow the elaboration of a priority score, and (iv) obtain the vulnerability measure of the patients on the surgical waiting list of the otorhinolaryngology unit. Using data analysis techniques, we propose strategic actions for ranking patients. This methodology could be useful for optimizing decision-making. The data allow the design of strategic actions for policy makers in the health care system. The data set can be used by other researchers to carry out studies in the area of patients on waiting lists.   degree of importance granted by the physicians to each of the variables. Finally, Table 8 list the diagnoses more frequently of otorhinolaryngology unit in Hospital of Talca.

Construction and completeness of the survey
We prepared, together with physicians, a survey that allows data collection to determine the score and vulnerability measures of patients on the surgical waiting list. The survey was taken in the polyclinic of the otorhinolaryngology unit of the Hospital of Talca for a period of five months, between February 05, 2018 and June 29, 2018. The process of starting the patient's file was carried out at first by a nurse and then finished by a physician. For construction of the dataset, five nurses and seven physicians participated in collecting data. In order to ensure accuracy and consistency of the records, each survey was reviewed by a nurse who verified the patient's medical history and health information system. The survey used can be seen in Fig. 1.
The complete instrument is the one specified in Fig. 1 and is the one used by the clinical team to complete the information. Besides the 20 parameters considered in this work, there are some entries (e.g., "belongs to program", "does the patient requires special equipment in surgery?") that are not relevant for our analysis since, as pointed out by physicians, they do not influence the score nor the vulnerability.
Once the survey was completed for each patient who entered the surgical waiting list, we used the individual files for tabulation and subsequent analysis. Then, the data was normalized and prepared for Table 1 Opinions of the physicians regarding priority by age group.

Types
Age group (years)  P1  P2  P3  P4  P5  P6  P7   Infant  0e1  1  3  5  6  6  5  6  Child  2e1 2  2  6  6  6  6  6  5  Teenager  13e1 8  3  6  4  6  6  4  4  Young adult  19e4 0  4  5  1  6  6  3  3  Adult  41e5 9  5  4  2  6  6  2  2  Elderly greater than 60 6 1 3 6 6 1 1 statistical analysis in the same unit of measurement to avoid scaling problems and some outliers were detected. Nevertheless, the raw data were considered as physicians reviewed and validated this information, as well as the specific characteristics of the patients. In Fig. 2, we show the structure of the normalized data distribution from 0 to 1 for each parameter, where we can have a notion of the mean and dispersion of the observations. Table 1 shows the opinion provided by each of the seven physicians when they were consulted about the priority in respect to age group. However, and in subsequent meetings to validate the  parameters and prioritization variables, the physicians decided to exclude this parameter since they considered it discriminatory; the same situation occurred with the gender parameter. In summary, we collected 205 surveys, 105 were women and 100 were men. Concerning the ages of the patients, 61.5% of the cases were between 0 and 20 years, 12.5% were between 21 and 40 years, 15.4% between 41 and 60 years, and 10.6% of cases were patients older than 60 years.

Raw data analysis process
Subsequently, to create a prioritization and vulnerability criteria of patients on a surgical waiting list, we processed the raw data and how each of the parameters and biopsychosocial variables that had been measured impacted the patients was discussed with the physicians. To do this, we performed the following steps. 1. We interviewed and consulted each physician looking for relevant parameters and variables that determine the priority of their patients. 2. Each physician quantified each parameter with a score between one and ten, where ten means that the parameter is crucial and a score of one means that the parameter is uninformative. 3. Additionally, physicians scored each parameter. 4. We consolidated raw data of all the opinions of the physicians for analysis. 5. We obtained the average of the opinions of each physician interviewed.
We conduct a review of the literature to know how to prioritize waiting patients in other parts of the world. After that, we met with the physicians and jointly defined the instrument. Then we take the records and perform only descriptive statistics of the data. Also, and for each parameter, we perform the measures of central tendency that can be seen in Table 2.

Description of parameters w i
For the experimental design of the survey, we explain in Table 3 Table 4 shows the opinion of the physicians and the value that each of them gave to the 20 parameters. Thus, in order to measure the level of importance of the parameters, they gave a score of 0 to the unimportant ones and up to 10 to the most important ones. To determine the total weight of each parameter, we divided the sum of each parameter's opinions in respect to the sum of all parameters. For example, the parameter Sever received 69 points (given by the opinions 10, 10, 10, 9,10, 10, 10, 10). Then, that value was divided by 853 points (which represented the sum of all the parameters). Then, the relative weight of Sever parameter was 8.1%. These weights can be seen in Table 4 and will be used as w i (%). Table 3 shows the values of the categorical parameters, which were mapped to numerical values to facilitate the calculation of the score. As a explain before, the w i parameter is common for all patients, parameter Sever becomes a variable a i;p , for example, Sever ¼ low. Table 5 shows that the parameter Sever was associated to a "low" severity with 7 points, a "medium" severity with 31 points, and a "high" severity with 70 points. Therefore, the relative score of Sever ¼ low is 7/108$100% ¼ 6.5%, of Sever ¼ medium is 31/108$100% ¼ 28.7% and of Sever ¼ high is 70/108$100% ¼ 64.8%. Now, for example, in order to obtain the contribution of Sever in the calculation of a patient's score s p with Sever ¼ high, the relative weight of the parameter Sever, which is 8.1%, is multiplied by the relative score associated to Sever ¼ high, which is 64.8%. The weights of the other variables are shown in Tables 5e7, and will be used as a i;p to indicate the importance of each of the elements of the parameter w i .

Score
The contribution of each i parameter to the patient score p is given by z i;p , which is obtained by multiplying the relative importance of the factor i, w i , and the value of said factor associated with the patient p, a i;p . Finally, we denote as s p to the patient's final score p, which corresponds to the sum of all the z i;p , i.e.,

Vulnerability
The objective of constructing a measure of vulnerability is to keep patients who might develop comorbidities or increased illness due to waiting more visible on the waiting list. In a similar way with [1], we will construct a way to measure the vulnerability of patients on surgical waiting lists, which we present below; v p;t ¼ f t À f p Jclin d;p;t (2) where v p;t represents the level of vulnerability of the patient p in the moment t, f t is the moment when the vulnerability is measured, f p is the patient's date of admission to the waiting list pand Jclin d;p;t corresponds to the physician's criteria d in relation to the patient's maximum waiting time p in the moment t. For more details, see [2].

Dynamic prioritization and patient choice
Together with the physicians, we have defined the criteria for dynamic prioritization and patient choice. Dynamic prioritization is built for each diagnosis in Table 8 and how they evolve, and in the patient choice, physicians select patients for surgery while simultaneously assessing their dynamic score and vulnerability measures. For more details of the methodology, see [2].
The proposal makes sense to physicians because the classification system allows them to keep a constant watch on the risk levels of waiting patients. It is for this reason that our proposal focuses on maintaining control of the list through the score and vulnerability of patients.
Technological Research, CONICYT, through the grant FONDECYT N.1180670 and through the Complex Engineering Systems Institute PIA/BASAL AFB180003. L. Gonz alez-Martínez is financed by the doctoral scholarship of the program Sistemas de Ingeniería of Universidad de Talca.