Original Article
The Minimum Basic Data Set (MBDS) as a tool for cancer epidemiological surveillance

https://doi.org/10.1016/j.ejim.2016.06.038Get rights and content

Highlights

  • Accuracy of the cancer incidence in general measured by MBDS has not been evaluated.

  • Minimum Basic Data Set (MBDS) has a sensitivity to detect cases of cancer above 60%.

  • MBDS has high specificity and NPV, and moderate NPV to detect cases of cancer.

  • The performance of MBDS in the detection of tumors depends on the type of cancer.

  • “Minimum Basic Data Set” could be a valuable tool in the monitoring of cancer.

Abstract

Objective

This work aims to evaluate the Minimum Basic Data Set (MBDS) as a data source in the detection of malignant tumors and explore its usefulness as a tool for epidemiological surveillance of cancer.

Materials and methods

MBDS hospital data discharge from Prince of Asturias University Hospital (HUPA, Alcalá de Henares, Madrid, Spain) and cancer cases recorded in the Hospital Cancer Registry (HCR) have been collected for the period between January 2012 and June 2014. Both databases have been linked by the number of clinical history. For the process of evaluation of MBDS, the types of cancer with more than 100 cases have been analyzed and sensitivity, specificity, positive and negative predictive values(PPV, NPV) of MBDS were calculated using as reference the diagnoses recorded in the HCR.

Results

3438 cases of cancer were accounted in the MBDS and 2445 in the HCR. The MBDS has a sensitivity to detect cases of cancer above 60%, although it varies depending on the type of tumor, reaching the highest values for bladder cancer. The specificity and the VPN were very high for all types of cancer studied, always on top of 95%. Finally, the VPP is generally moderate, between 50% and 70%.

Conclusions

The systematic exploitation of the MBDS can provide a valuable tool in the monitoring of cancer by its acceptable sensitivity and high specificity, allowing obtaining information without the delays involved in the consolidation of the annotations of the HCR. Furthermore, its use could partly mitigate the lack of data in important regions of Spain.

Introduction

The epidemiological surveillance of cancer is based on the monitoring of time trends and geographical distribution of its indicators of incidence, mortality and survival. This monitoring aims to assess the burden of cancer by generating alerts when increases in indicators or strong inequalities are detected, requiring intervention in public health. Data from these surveillance systems also provide information for rational use of resources and allows the impact of cancer prevention programs evaluation.

Mortality monitoring use death certificates as a primary data source. In Spain death certificates are managed by the local counties' mortality registries and subsequently centralized by the National Institute of Statistics (INE). This system has a universal coverage.

For the study of the incidence and survival, population-based cancer registries are needed. Another source of data in new cases is the hospital cancer registries (HCR). The majority of cancer patients need to be admitted to a hospital and this type of records exists in many centers.

The admission of a patient with cancer by a health center leaves an administrative ‘footprint’ through the Minimum Basic Data Set (MBDS). The MBDS has state coverage, mandatory completion and linked administrative data with diagnoses. Therefore, it represents a useful additional source of data to feed hospital and population-based cancer registries, and it can be a valuable tool in epidemiological surveillance of cancer.

One of the features that points to the MBDS as an effective instrument for surveillance of cancer, is the immediacy in the availability of the collected data. Thus, while the delay in the availability of population-based registers of cancer data can be of several years and in the HCR is rarely less than one year, access to data from the MBDS is almost immediate. The 15th of each month, hospitals send coding of all diagnoses of the hospitalized patients discharged during the previous month to its corresponding department of health, in each local county.

MBDS has been used in Spain to numerous studies of epidemiological surveillance of various infectious diseases such as salmonella and syphilis [1], [2], varicella [3], or psychiatric illnesses such as depression [4]. In the case of cancer it has been used to study the extent of admissions for cervical cancer during the period 1999–2002 [5]. However, the accuracy of the cancer incidence measured by the MBDS has not been evaluated in Spain for cancer in general, there are only some pilot studies for prostate [6] and colorectal cancer [7].

This work aims to evaluate the MBDS as a source of data in the detection of malignant tumors and explore its usefulness as a tool for cancer epidemiological surveillance (CES).

Section snippets

Material and methods

To achieve the objectives described above, MBDS hospital data discharge from Prince of Asturias University Hospital (HUPA, Alcalá de Henares, Madrid, Spain) and cancer cases recorded in the Hospital Cancer Registry (HCR) have been collected for the time period between January 2012 and June 2014.

The HUPA HCR started its activity in 2008 based only on pathology reports, and the MBDS was added as source data in mid-2011. During the study period, data sources that feed the HUPA's HCR were: MBDS,

Results

For the study period, the HCR contains 2492 tumors cases corresponding to 2445 patients (MRNs) and the MBDS contains 32,148 hospital admissions corresponding to 31,058 MRNs. Of these, 3357 have a diagnosis of cancer. Table 2 shows the number of the registers in both data sources between 2012 and 2014 that have a diagnostic of those types of cancer with 100 or more records in the MBDS. There are more cases of cancer in the MBDS than in the HCR, regardless of analyzed tumor subtype. On the other

Discussion

According to the results of this study, the MBDS has a sensitivity to detect cancer that varies depending on the type of tumor, usually above 60%, reaching the highest values in urinary bladder cancer. With regard to the specificity and the NPV, the MBDS presents very high values in all types of cancer studied, always between 95% and 100%. On the other hand, the PPV of MBDS is generally moderate, between 50% and 70%, although it also depends on the group of tumors studied.

Current resources for

Conflict of interest

The authors declare there is no conflict of interest.

Acknowledgments

We thank Virginia Mehdi Hueros 1 for her English review of the manuscript.

References (13)

There are more references available in the full text version of this article.

Cited by (16)

View all citing articles on Scopus
View full text