Data on Vietnamese patients׳ behavior in using information sources, perceived data sufficiency and (non)optimal choice of health care provider

This data article introduces a data set containing 1459 observations that can enable researchers to examine issues related to and perform statistical investigations into questions of relationships between sources of health care information, data sufficiency, trust levels between patients and healthcare experts (and the advice). The data set also records assessment of Vietnamese patients on whether their choice of health care provider is best available (optimal vs. nonoptimal). The data come from a survey in many hospitals in Hanoi and several neighboring provinces/cities in the North of Vietnam, during the last quarter of 2015. Variables that can be useful for future analysis include sources and availability of information, cost, and amount of time for seeking information. The quality of information and health professionals’ credibility are critical factors in helping patients choose a health care provider. Mendeley Data, v1 http://dx.doi.org/10.17632/gmbz53tpwc.1; and can enable the modeling after useful discrete data models such as BCL, with one example being provided in this data article.


a b s t r a c t
This data article introduces a data set containing 1459 observations that can enable researchers to examine issues related to and perform statistical investigations into questions of relationships between sources of health care information, data sufficiency, trust levels between patients and healthcare experts (and the advice). The data set also records assessment of Vietnamese patients on whether their choice of health care provider is best available (optimal vs. nonoptimal). The data come from a survey in many hospitals in Hanoi and several neighboring provinces/cities in the North of Vietnam, during the last quarter of 2015. Variables that can be useful for future analysis include sources and availability of information, cost, and amount of time for seeking information. The quality of information and health professionals' credibility are critical factors in helping patients choose a health care provider.
Mendeley Data, v1 http://dx.doi.org/10.17632/gmbz53tpwc.1; and can enable the modeling after useful discrete data models such as BCL, with one example being provided in this data article. &

Data accessibility
Datasets are provided with this article.

Value of the data
The data help acquire understanding about patients' demand for health information before choosing health care provider.
Assessments of patients access to different sources of information and data, and values in their decision making process.
The data enable researchers' further examination into alternative functions of available but seemly underutilized public information system and health service such as the public emergency medical service hot line 115.
The data potentially offer an opportunity of examining the quality of medical information from different sources and perception of efficiency in Vietnamese patients' choice of health care provider.

Data
The data set contains 1459 records obtained from a survey of assessments from Vietnamese patients about information sources, time consumption and labor cost for acquiring information, the perceived value of information and efficiency in choice of health care provider.
The age distribution of patients participating in the survey is in Fig. 1. Discrete (categorical) variables are measured and reported in the survey data set (see Table 1).

Experimental design, materials and methods
The data can be employed by the multi-category logit models to enable analysis based on baselinecategory logits (BCL), for computing probabilities upon events of hypothetical influence. The logic for designing the experiment and thus data set is described as follows. A patient (among n) is treated as independent and identical. Each data point has outcome in any of J categories for each factor to be investigated. Let y ij ¼ 1 if patient i has outcome in category j, and y ij ¼ 0 otherwise. Then, y ij ¼ y i1 ; y i2 ; …; y ic À Á represents a multinomial trial, with P j y ij ¼ 1. As n j ¼ P j y ij the number of trials having outcome in category j, the design is based on the assumption that n 1 ; denote the probability of outcome in category j for each patient, the multinomial probability mass function is multinomial with corresponding sets of probabilities π 1 x ð Þ; …; π j x ð Þ È É . Thus, each response is aligned with a baseline category.
BCL models measure the effects of x (J-1) logits, which in general vary according to the response paired with the baseline category, providing for parameters for these logits.
The empirical dataset will then be used to evaluate Pearson-type likelihood ratio test statistics (X 2 ; G 2 ) for goodness-of-fit, following a multivariate generalized linear model (GLM) estimations. Technical details for practically estimating multinomial logistic models is provided in Ref. [2]. Applied analysis can be performed in R (see [3]). Practical uses of survey data can be referred to Ref. [4].
Some possible questions and hypotheses worth testing of, using the data set [1], is in Table 2.
One example of the analysis is to compute response probabilities from multinomial logits, i.e., with P j π j x ð Þ ¼ 1; α J ¼ 0 and β J ¼ 0. In the following  The residency status of a patient Resident (res), non-resident from other urban areas (nonres.urb), from a rural area in the northern rivers delta regions (rurdelta), remote areas, e.g., mountainous regions (remarea) Table 2 Possible research questions arising from the data set.
What are the effects of accessibility to information (through various sources: friends/relatives, mass mediawith a focus on the Internet,and health care experts) on patients' perception of information sufficiency when having to make a choice regarding a health care provider? How are these sources of information different in terms of their influence on patients' perception? What are the measured effects of time and costs spent by patients on ex ante probabilities of acquiring sufficient information for decision-making? What are the effects of socioeconomic status (SES) and residency status on data/information sufficiency for patients' decision making? Are the ex post probabilities of making an optimal decision conditional upon accessibility to expert information regarding health care and the level of trust in the expertize provided? Is the effect of mass media/Internet use significant? In what ways do the costliness of information and trust in expertize affect the outcome of a patient's choice? Are the use of 115 Emergency Hot-line counseling and the status of residency having significant impacts on patients' choice outcomes (optimal vs. nonoptimal impacts)? Table 3 Patients' perception regarding information sufficiency following their access to experts and friends/relatives. "x11.convrel" " x12.convexp" " x43.info" "Sufficient" " Insuff" "low.convrel" " low.convexp" 27 99 "med.convexp" 8 2 5 "hi.convexp" 9 6 "med.convrel" " low.convexp" 67 164 "med.convexp" 112 169 "hi.convexp" 58 23 "hi.convrel" " low.convexp" 125 123 "med.convexp" 109 108 "hi.convexp" 162 65 Table 4a Distribution of patients who rely on information from friends/relatives and mass media/Internet sources, with respect to data sufficiency.
convrel"; and "x13.convint": "hi.convint". Residual deviance: 25.45 on 4 degrees of freedom Table 8 Empirical probabilities of data sufficiency following access to friends/relatives and mass media/Internet sources. "x43.info" "Sufficient" "Insufficient" "x11.convrel"| "x13.convint" "low.convint" "med.convint" "hi.convint" "low.convint" "med.convint" "hi.  Table 6 shown below reports the full empirical distributions of probabilities over different categorical values of factors "x12.convexp" and "x11.convrel." As a familiar practice, when facing difficulty in accessing expert counseling, Vietnamese patients choose to consult with family members and close friends. Likewise, the estimated coefficients from multinomial logistic regression with independent variables are "x11.convrel," "x13.convint" and the dependent variable is: In this example, computed probabilities show the effects of both information from friends/relatives and from mass media/Internet on patients' data sufficiency. Such empirical probabilities are provided in Table 8, using the relationships established in the estimated coefficients of Table 7. Fig. 2 below is drawn using computed values in Tables 7 and 8 with respect to the changing sociocultural value in the society [5].
The changing shapes of the graphs in Fig. 3 show that the positive effect of expert counseling is stronger than that of mass media/Internet, and friends/relatives information source is critically important.