According to the American Cancer Society and the National Cancer Institute’s cancer registries, cancer is the second leading cause of death in the state of Texas. However, in many rural communities where cancer treatment facilities are not located, the cancer incidence statistics are suppressed to prevent duplication of counts from the reporting county where the patient is treated. An overview of the cancer incidence and trends of the top ten cancers for the United States and all sites combined for the rural area of Culberson County, Texas was performed for the years 2000 through 2009.
Data for the study was obtained based on reports from the International Classification of Diseases (ICD-9) and medical records from Culberson County Hospital and the Van Horn Rural Health Clinic in Culberson County, Texas. A total of 429 actual new cancers were identified over the ten year period. This is an average of 42. 9 official new cases each year for a rural population of only 2431. After adjusting these numbers to make comparable at a “per 100,000” population, the cancer incidence rate for this target area was found to be three times the national and state average.
With medical advancements in cancer treatment and the aging of the population, the cancer incidence will likely continue to rise for years or even decades to come. Accurately accounting for rural areas will provide information on the impact of cancer on the health care system, document the need for increased research, and further the progress in supporting cancer control knowledge across all segments of the population. In the state of Texas, cancer is the second leading cause of death (Wilkerson and Smith, 1995).
The incidence of various cancers in Culberson County (population 2,431), which has it’s primarily residence as the town of Van Horn, Texas (population of 2,013), is largely unknown (www. city-data. com). Most patients who present to Culberson County Hospital and the Van Horn Rural Health Clinic are referred to other facilities, and as such do not get counted for this county. The possibility of developing cancer can have an effect not only on individuals but also on the medical management of patients undergoing routine clinical care as well.
In an attempt to better quantify the numbers of cancer incidence in the county, the records of Culberson County Hospital and the Van Horn Rural Health Clinic were examined over a ten year period of 2000 through 2009 to investigate the rate of cancer incidence in Culberson County. This preliminary study was prompted due to a strong recent opinion among various Culberson County Hospital providers in Van Horn, Texas. There is the perception that local healthcare providers are seeing many more patients with cancer symptoms than previously (Friday, A. Md. , Pers. Comm. ).
As Culberson County Hospital is a frontier hospital, it does not have the staff or resources to treat a range of specialty conditions, and thus refers patients to El Paso, Midland, Odessa, Lubbock, Dallas, Houston and elsewhere. The patients being referred to oncologists outside of Culberson County for official diagnosis are not reported in their county of residence but in the county of treatment. This is due to the method of reporting by the referred oncologist.
In an effort to prevent duplication of state statistics and inflate occurrence percentages, reports from the home provider are omitted (Carrillo, J. , Pers. Comm. ). However, this still potentially results in artificial increases in counties with cancer treatment facilities, and subsequent decreases in areas that don’t. As a result, there is little data to validate these notions. The number of cancer cases, especially in rural areas, has an effect on the demand placed on the local health care system.
The key to survival for cancer is prevention and early detection, which is influenced by the type of cancer prevalent in an area, the extent of the disease, and treatment options. Quality tracking of cancer cases provides vital information to communities to assist in early detection. Hence, this study was undertaken to investigate the hypothesis of a potential increasing trend in the number of cancer incidences in comparison to statistics based on the Texas Bureau of Vital Statistics (http://www. cancer-rates. info/tx/), National Cancer Institute (http://seer. ancer. gov) and Texas Department of Health (http://www. dshs. state. tx. us/).
The Surveillance, Epidemiology, and End Results (SEER) program has a complete source of population-based data in the U. S. that includes stage of cancer at the time of diagnosis and follow-up of all patients for survival data (Howe et al. , 2006). Although records are amassed from health care providers, hospitals, laboratories, nursing homes and treatment facilities, only those in the registry’s defined geographic area are documented (Howe et al. , 2006).
By 1975, SEER was reporting data each November to the National Cancer Institute (NCI) (Howe et al. , 2006). The most recent reports for cancer trends are based on the diagnosis years of 1975–2007. Furthermore, each state registry collects data on patient demographics and primary cancer site, which is also submitted to NCI for U. S. statistics. Despite that cancer registries continuously update their databases, the acceptance of various records, and the disallowance of other records that would cause duplication of statistics, creates many reporting delays (Howe et al. , 2006).
It is the hope of this study that such reporting delays, in addition to the confusion with patient county demographics, can be discussed. Methods –Prior to initiating any medical record review, HIPAA (Health Insurance Portability and Accountability Act (HIPAA of 1996) disclosure forms were signed. Maintaining confidentiality of medical information is mandated by law and patient privacy was held at highest priority during this study. HIPAA regulations were assessed and conducted this study in compliance with proper protocol for research data gathering (Anzaldua and Sanchez, 2002).
Review of records in preparation for research – Under the “preparatory to research” provision, covered entities may use or disclose PHI to researchers to aid in study recruitment. The covered entity may allow a researcher, either within or outside the covered entity, to identify, but not contact, potential study participants under the “preparatory to research” provision. However, before permitting this activity, a covered entity must receive proper representation that (1) the use or disclosure is sought solely to review PHI for activities preparatory to research, and (2) that the PHI sought is necessary for the research purposes.
A researcher may contact potential study participants without an Authorization from the individual, if the researcher is a workforce member of a covered entity or a business associate of the covered entity (Gates, 2003). Data collection began with reviewing monthly International Classification of Diseases (ICD-9-CM) reports from Culberson Hospital which is reported to the Texas Department of Health – Cancer Registry. The information for these reports is gathered from local provider referrals, biopsies, and lab reports for each month. Excluded from the research were death certificates.
Due to privacy issues and HIPAA regulations, all medical records were retained at Culberson Hospital. For each patient record, a dedicated coding system for disease identification is indicated. The principal manual for classifying diseases used by healthcare providers is the International Classification of Diseases (ICD), published by the World Health Organization. Various cancers are coded according to the International Classification of Diseases, 9th edition (ICD-9) (Buck, 2004) or the more specialized International Classification of Diseases for Oncology, 3rd edition (ICD-O-3) (Percy et al. 2000).
The ICD-9 reports were used for this study due to their dominate usage as a diagnostic classification based on the multilevel arrangement of site, morphology, behavior, and grading of neoplasms that system categorizes malignant neoplasms for all types of tumors. ICD is used principally in cancer registries for coding the site (topography) and the histology (morphology) of neoplasms, usually obtained from a pathology report (Muir and Percy, 1991). A neoplasm is an abnormal new growth of cells or tissue that will continue to grow if not treated and become a tumor (McKean, 2005).
The major advantage of using the ICD for research purposes is that the ICD-9 is globally accepted and has been published by the World Health Organization (WHO) since its 6th revision in 1946 (Jemal et al. , 2007). Additionally, ICD is used by all WHO Member States for most health statistics (Jemal et al. , 2007). The ninth revision, in which codes are valid from October 1, 2009 through September 30, 2010, is the current edition in use at the time of this study. Most classifications assign numerical codes so that recurring information can be expressed, stored and retrieved easily.
The ICD-9 uses codes ranging from 140. 0 to 796. 76 for diagnosis and V10. 0 through V87. 41 for code explanation. This information conveys three levels of description: malignancy, organ affected, and histological type. ICD-O-3 codes add a separate one-digit code for histologic grading (differentiation) (Hannah and Ball, 2003). Neoplasms can be categorized in many ways and involves allocation into classes according to the ICD system. The two most important factors for proper classification of neoplasms for clinician and cancer registry are the location of the tumor in the body and the morphology, (i. . , the form of the tumor under microscopic examination) (dos Santos Silva, 1999). This indicates its behavior (malignant, benign, in situ, and uncertain).
Cancer registries attempt to classify each neoplasm according to its topography, morphology and behavior, as well as recording particulars of the host (Muir and Percy, 1999). The statistics presented, with the exception of mortality, were obtained from review of records from both Culberson Hospital and Van Horn Rural Health Clinic, as these records are managed separately.
Over sixteen-hundred different codes were pulled from the ICD-9 reports, for which each patient’s medical record was retrospectively searched to verify a referral to an oncologist, and/or reported an abnormal screening. A binder was created for each year for quick reference as needed and maintained solely at Culberson Hospital. Based on the local physician’s referral, follow up with respective oncologists was conducted for confirmation of diagnosis and official diagnostic record that would link the patient back to Culberson County. All patients’ records were verified as being Culberson County residents prior to diagnosis follow up.
Once all ten years were reviewed, a patient list was compiled and sent to each referred oncologist for confirmation with the provider code. Additionally, personal contact was made with Texas Oncology in El Paso, Texas, for verifying the accuracy of confirmed cancer treatments. In retrieving data, it was necessary to convert some descriptions and codes in order to group into broader classifications for certain sites, and each specific diagnostic code had to be matched with a more general cancer site in order to make local statistics comparable to state and national statistics (Doll and Peto, 1981).
Another issue that arouse during this study was carefully identifying any codes that had been changed over the years with the revisions of the ICD. Data was converted from a detailed to a more generalized version, but the impact of this change of content was very difficult to assess. In converting from one revision to another, many terms listed were only in the alphabetical index. As different editions are sometimes indexed differently from one another, this created additional complications.
These changes made it more challenging to compare data over long periods of time (Hayat et al. , 2007). A final spreadsheet was created categorizing each diagnosis for each year reviewed. Each worksheet was compared from previous years to insure only primary cancers and initial secondary cancers were counted. Duplications were highlighted and eliminated manually from counts. This spreadsheet was analyzed statistically using Microsoft Excel (Redmond, USA) and generated corresponding tables and graphs of the outcomes (dos Santos Silva, 1999; Zar, 2010).
The incidence rates presented in this study are expressed as the number of new primary cancers per year as well as for second or multiple cancers occurring among survivors previously diagnosed with cancer. To accurately compare to state and national statistics, the results for each site and year were multiplied by 41. 13 in order to modify the actual population of 2431 to a target population that could be expressed in the statistical rate of “per 100,000” people. Data analyzed was approximately 98% complete for all cancer sites combined, as some cancer sites, such as melanoma, were less complete in the records.
We recognize that some residents of Culberson return to their home provider for annual checkups and serious illness, and there are possible cancer patients not originally seen by a Culberson County provider that would not be accounted for. Therefore, the results of this study are to be considered conservative. While it is understood that these results will not be complete, it is believed that the above methodologies will produce the most accurate data for cancer occurrence in Culberson County to date.
Results –This study presents the cancer occurrence of new cases among residents of Culberson County from 2000 through 2009. Examination of over 48,000 documents over the ten year period yielded a total of 429 new cancers in Culberson County, Texas. This is an average of 42. 9 new cases each year. The actual number of new cancer cases for each cancer sites are shown in Table 1. When rates are adjusted to the state and national population for comparisons, these results report cancer rates approximately three times the state and national average (Figure 1) (Anzaldua and Sanchez, 2002).
Included in these cancer totals were 255 men (59%) and 174 women (41%), with ages ranging from 3-101. The ten most commonly diagnosed cancers in the United States and Texas were used for comparisons with statistics in Culberson County (Anzaldua and Sanchez, 2002). These top ten sites include prostate, breast, lung and bronchus, colon and rectum, corpus and uterus and non-specific, urinary bladder, non-Hodgkin lymphoma, melanomas of the skin, kidney and renal pelvis, and cancer of the ovary (Figure 2) (www. cdc. gov).
These ten sites accounted for 59% of total cancers in men in Culberson County and 41% of total cancers in women (Anzaldua and Sanchez, 2002). Within these 429 patients, an average of 9. 2 cases per year (41%) of all cancer sites were diagnosed as melanoma, and prostate cancer was the leading type of cancer diagnosed in men at an average of 4. 1 cases per year (18%) (Figure 3).
Melanoma occurrence separated by gender resulted in 62% occurrence in males, while 38% of females developed this cancer with the median age for both being 63. years of age (Anzaldua and Sanchez, 2002). Cancers of the lung and bronchus and cancers of the uterine were tied for the third most commonly diagnosed cancers in Culberson County. These four cancers together (melanoma, prostate, lung and bronchus, and uterine) accounted for 79% of all cancers reported among Culberson County residents during 2000 through 2009. Of additional interest, breast cancer accounted for 9% of all cases reported, with a total of 21 cases over the ten years within the study area.