Opinion
Opinion By: Gregory D. Stumbo, Attorney General; Amye L. Bensenhaver, Assistant Attorney General
Open Records Decision
The question presented in this appeal is whether the University of Louisville violated the Open Records Act in the disposition of Dr. James W. Coleman's February 19, 2004 request for access to documents containing information compiled and maintained by the University as a participant in the Kentucky Cancer Registry. For the reasons that follow, we find that the University's disposition of Dr. Coleman's request was procedurally deficient. Moreover, we find that although the University was actuated by its commitment to strictly observe the privilege relating to Registry "information, interviews, reports, statements, memoranda, or other data," set forth at KRS 214.556(6), its reliance on that provision, incorporated into the Open Records Act by operation of KRS 61.878(1)(l), to withhold the statistical data sought was misplaced.
Following an earlier, unsuccessful attempt to gain access to, inter alia, "raw data and any and all other documents used in connection with the Breast Cancer Incidence calculations that appeared in a report published by the U of L School of Public Health and Information Services in September 2003," Dr. Coleman submitted a revised request on February 19, 2004. The records to which Dr. Coleman requested access were identified as follows:
1. [A]ny document(s) showing the total number of breast cancer cases for each zip code that were used in connection with the Breast Cancer Incidence calculations in the Cancer Incidence in Jefferson County, Kentucky 1996-2000, a Report from the Louisville Kentucky Metro Health Department and the U of L School of Public Health and Information Sciences, Second Full Draft: September 4, 2003 ("Report").
2. [A]ny document(s) showing the total number of female breast cancer deaths each year for 1996, 1997, 1998, 1999 and 2000 that were used in connection with the Female Breast Cancer Mortality rate calculations that are summarized in the graphic Figure on page 2 of the Report.
3. [A]ny document(s) showing the numerical figures from the results of the age-adjusted calculations (to the US Standard 2000) for each zip code in Jefferson County, Kentucky for the female breast cancer incidence rates covering the period 1996 through 2000 . . . .
4. [A]ny document(s) showing the numerical figures from the results of the age-adjusted calculations for each zip code in Jefferson County, Kentucky for the "African American Breast Cancer" covering the period 1996 through 2000. These would be the selfsame, age-adjusted calculations used to aid in constructing the U of L document identified as "African American Breast Cancer" designed to show the distribution of the incidence rates by zip codes . . . .
5. Please provide access to any document(s) showing the total number of breast cancer cases for the African American calculations referred to in numbered Paragraph 4, above, i.e., by zip codes for 1996-2000.
7. Please provide access to any document(s) showing the total number of female breast cancer cases for each zip code in Jefferson County, Kentucky used to make the age-adjusted, female breast cancer calculations as set forth in a copy of the U of L document that is already in the public domain. See Attachment # 1.
With reference to each of these requests, Dr. Coleman proposed an alternative mode of access whereby the University would extract the information sought and create a record containing that information. Having received no written response to these requests, Dr. Coleman initiated an open records appeal to this office on March 12, 2004. 1
By letter dated March 23, 2004, Dr. William J. Morison, Open Records Officer, responded to Dr. Coleman's appeal on behalf of the University. 2 He explained that he believed Dr. Coleman's later request mirrored his earlier request and that he therefore did not give it first priority. Nevertheless, Dr. Morison indicated, he and Dr. Coleman had communicated by telephone and he had again invoked KRS 214.556 to generally deny the February 19 request. Working with Dr. Timothy Aldrich, Department of Epidemiology and Information Sciences, School of Public Health and Information Sciences, and an authorized user of Kentucky Cancer Registry data, Dr. Morison identified and produced for Dr. Coleman's inspection "responsive records in the possession of the University of Louisville that are not shielded by KRS 214.556," 3 but continued to rely on the privilege to deny him access to the records specifically identified in his request.
Thereafter, the parties to this appeal exchanged correspondence relating to the requests. The University maintained generally that if the requested records did not contain personally identifiable data, they would be released to Dr. Coleman. Dr. Coleman responded that the University had not fulfilled its obligation to provide "a written response stating whether or not [the University would afford him] access to the documents related to . . . requests in numbered paragraphs 1, 2, 3, 4, 5, . . . and 7." To facilitate our review, the Attorney General propounded two sets of questions to the University, pursuant to KRS 61.880(2)(c), 4 aimed at resolving the fundamental questions on appeal: Whether disclosure of the records to which Dr. Coleman requested access reveal the identities of persons whose condition or treatment have been reported to the Kentucky Cancer Registry in contravention of KRS 214.556(5), and the corollary question whether the privilege for "all information, interviews, reports, statements, memoranda or other data furnished by reason of KRS 214.556, and any findings or conclusions resulting from those studies . . ." extends to the records sought. KRS 214.556(6).
At the risk of oversimplification, we note that in response to our questions the University confirmed disclosure of "aggregate data values, e.g., Zip code level, age-adjusted incidence rates . . . as shaded regions on a map, e.g., colored for a range of rates," but resisted disclosure of "Zip code-specific rates . . . [or] case counts, stratified with details of age, race [and] gender [because they] were deemed to pose a violation of individual patient confidentiality." We further note, again at the risk of oversimplification, that Dr. Coleman referenced numerous publications of the Kentucky Cancer Registry and the University of Louisville containing like information for male and female breast cancer incidence rates by zip code, as well as an October 2003 email in which Dr. Aldrich referred to a November 2003 meeting he would attend, a meeting the University neither confirms nor denies, in anticipation of which he indicated he would be compiling "female only breast cancer rates . . . ." Commenting that in the period under scrutiny, 1996 to 2000, the Registry reported that there were "2,847 cases of non-invasive female breast cancer and 429 cases of invasive breast cancer (or a total of 3,276 cases in Jefferson County, Kentucky," and that there are 34 zip codes in Jefferson County, Dr. Coleman questioned the mathematical probability of tracing the identity of a person or persons based on the scant information requested, namely, gender, race, and zip code. The University responded that "a table cell with, for example, one African American female, age 30-34, diagnosed with breast cancer in 1998, in Zip code 402 - -," which it characterized as the "level of data Dr. Coleman seeks," is beyond its authority to release because "someone could potentially [be] recognized by the specific description . . . ." While we commend the University for its commitment to protect the identity of persons whose condition or treatment has been reported to the Kentucky Cancer Registry, consistent with the statutory mandate set forth at KRS 214.556(5), we find that a close review of Dr. Coleman's February 19 requests does not support the University's argument that disclosure of the records sought will breach the privilege codified at KRS 214.556(6). 5
In the single reported decision dealing with access to statistical information in the context of an open records appeal, the Kentucky Supreme Court prefaced its analysis with the following language:
The Open Records Act, KRS 61.871, states in pertinent part:
The public agency that is the subject of an Open Records request, has the burden of proving that the document sought fits with an exception to the Open Records Act. KRS 61.882(3) and University of Kentucky v. Courier-Journal, Ky., 830 S.W.2d 373 (1992). The statute demonstrates a general bias favoring disclosure. See Kentucky Board of Examiners of Psychologists v. Courier-Journal, Ky., 826 S.W.2d 324 (1992).
Hardin County Schools v. Foster, Ky., 40 S.W.3d 865, 868 (2001). Turning to the question of law before it, the Court concluded that the Hardin County Schools improperly denied a newspaper's request for statistical data pertaining to disciplinary action, including the offense prompting each action and the particular school in which the offense occurred. Citing KRS 61.878(2), which provides that "no exemption [in KRS 61.878(1)(a) through (1)] shall be construed to prohibit disclosure of statistical information not descriptive of any readily identifiable person," the Court concluded that "the statistical compilation sought . . . does not identify individual students and therefore is not descriptive of any readily identifiable person contemplated by the statute." Id. Amplifying on this view, the Court reasoned:
[T]he information sought by . . . the newspaper does not identify the names of any student nor reveal personal characteristics. The identity of the school, year of occurrence, reason for the disciplinary action and the type of action does not directly relate to any particular student. Personally identifiable information would include information that makes the identity of the student easily traceable, such as a name, address or personal characteristics.
Id. at 869. In sum, the Court opined, "the newspaper requested a statistical compilation, not individual records," and the school system improperly denied the newspaper access to that compilation. Id.
We are guided by the Supreme Court's opinion in Hardin County Schools v. Foster in our resolution of this appeal. Both involve agency reliance on a separate statutory enactment extending protection to a category of records and access to statistical compilations for which the protection is claimed. Fundamental to both is the question of whether disclosure of the statistical complication would enable the public to readily identify persons to whom the protection extends. As in Hardin County Schools v. Foster, we find that that question must be answered in the negative.
The statutory enactment upon which the University relies is codified at KRS 214.556, establishing the Kentucky Cancer Registry. The pertinent portions of that statute provide:
(5) The identity of any person whose condition or treatment has been reported to the Kentucky Cancer Registry shall be confidential, except that:
(6) All information, interviews, reports, statements, memoranda, or other data furnished by reason of this section and any findings or conclusions resulting from those studies shall be privileged.
It is unclear why the General Assembly elected to employ the term "privileged" in subsection (6) of KRS 214.556, given the narrower legal meaning generally attached to that term. 6 It is clear, however, that subsections (5) and (6) operate in tandem to prohibit disclosure of personally identifiable information flowing to or from the Kentucky Cancer Registry in furtherance of its stated purpose "of providing accurate and up-to-date information about cancer in Kentucky and facilitating the evaluation and improvement of cancer prevention, screening, diagnosis, therapy, rehabilitation, and community care activities for citizens of the Commonwealth." KRS 214.556(1). Pursuant to this statutory charge, the Kentucky Cancer Registry periodically reports its data and related findings and recommendations to designated legislative and executive agencies, KRS 214.556(7), and issues publications such as the "1996-2000 Cancer Incidence Report, Publication Year 2002" to which Dr. Coleman refers. So too, the University of Louisville disseminates the results of its studies in publications such as "Cancer Incidence in Jefferson County 1996-2000." 7 The prohibition on disclosure of data is not, therefore, absolute, but is instead confined to data identifying "any person whose condition or treatment has been reported to the Kentucky Cancer Registry" per KRS 214.556(2). 8
Based on the Supreme Court's analysis in Hardin County Schools. v. Foster, above, we conclude that the University has failed to meet its burden of proving that the statistical information contained in the records to which Dr. Coleman requests access is descriptive of any readily identifiable person. To begin, the University misstates the statistics sought. Our review of his February 19 requests confirms that the data he seeks is limited to the total number of breast cancer cases for each zip code for the period 1996-2000; the total number of female breast cancer deaths for the same period; the total number of female breast cancer cases for each zip code for the same period; and the total number of African American breast cases for each zip code for the same period. Dr. Coleman did not, for example, request "case level data" or "case counts" containing personal identifiers such as age, much less unique personal identifiers such as name, social security number, and address. While we understand and appreciate the University's reluctance to disclose statistical information that could be used to identify a particular person, we believe that the statistical information Mr. Coleman seeks is, as in Hardin County Schools v. Foster, above, not descriptive of any readily identifiable person. For purposes of comparison, we find that the identity of the school in Hardin County Schools can be analogized to the zip codes at issue in this appeal, 9 that the year of occurrence in Hardin County Schools can be analogized to the year of diagnosis in this appeal, and that the reason for the disciplinary action and type of action taken in Hardin County Schools can be analogized to the type of cancer diagnosed and the race and gender of the person whose condition was diagnosed. We acknowledge that the latter are "personal characteristics," but are not persuaded that they are descriptive of any readily identifiable person. As in Hardin County Schools v. Foster, above, at 868, "The statistical compilation sought . . . does not identify individual [persons]," and the University has not met the burden of proving that those records fit within KRS 214.556(5) and (6). 10 And, as in Hardin County Schools v. Foster, the General Assembly has recognized the strong public policy in advancing cancer control, and mandated the collection and analysis of data in support of that policy. Disclosure of statistical compilations not descriptive of any readily identifiable person promotes that policy as well as the stated policy of the Open Records Act. Consistent with this decision, the University may elect to produce redacted copies of the records specifically requested, or generate records responsive to Dr. Coleman's alternative requests.
In closing, we note that the University violated KRS 61.880(1) in failing arHHarHto respond in writing, and within three business days, to Dr. Coleman's request. We are confident that the University is fully aware of its legal obligations under the cited provision, and will not unnecessarily lengthen this decision with a recitation of those obligations. Nevertheless, we urge the University to rededicate its efforts to strict compliance with these legal obligations notwithstanding the press of other open records business.
A party aggrieved by this decision may appeal it by initiating action in the appropriate circuit court pursuant to KRS 61.880(5) and KRS 61.882. Pursuant to KRS 61.880(3), the Attorney General should be notified of any action in circuit court, but should not be named as a party in that action or in any subsequent proceeding.
Dr. James W. Coleman8622 Blackpool DriveLouisville, KY 40222-5667
Dr. William MorisonEkstrom Library, Room 421Louisville, KY 40292
Angela KoshewaOffice of General CounselGrawemeyer Hall, Room 206University of LouisvilleLouisville, KY 40292
Footnotes
Footnotes
1 Dr. Coleman requested access to additional documentation consisting of color-code legends for published breast cancer incidence maps, and mathematical formula and legends for symbols and characters used in formulating age adjusted calculation results and age-adjusted female breast cancer mortality rates. The University belatedly furnished Dr. Coleman with these records thereby mooting this portion of his appeal. 40 KAR 1:030 Section 6.
2 Coincidently, Dr. Morison had contacted Dr. Coleman by letter dated March 16 to confirm their earlier telephone conversations relating to the February 19 request and to explain that he was then dealing with "a flood of open records requests." Dr. Morison advised that Dr. Coleman's request had been referred to the "appropriate U of L official" with instructions to respond no later than March 19 "concerning the existence of any university records that might be responsive to [his] request."
3 The records produced consisted of various reports, and portions of treatises and articles including: 1) "Cancer Incidence in Jefferson County 1996-2000: A Report"; 2) "Central Cancer Registries: Design, Management, and Use"; 3) "Epidemiology"; 4) "Concepts of Epidemiology"; 5) "Statistical Methods for Rates and Proportions." Dr. Coleman characterized these records as "gratuitous" and nonresponsive.
4 KRS 61.880(2)(c) provides as follows:
On the day that the Attorney General renders his decision, he shall mail a copy to the agency and a copy to the person who requested the record in question. The burden of proof in sustaining the action shall rest with the agency, and the Attorney General may request additional documentation from the agency for substantiation. The Attorney General may also request a copy of the records involved but they shall not be disclosed.
(Emphasis added.)
5 In the course of this appeal, much discussion has been devoted to the question of whether responsive records currently exist. It is only incumbent on the University to produce records that have already been created, but this includes records and report queries utilized in the development of the referenced publications and reports, as well as associated studies. Pursuant to KRS 61.878(4), the University may, of course, redact any information not specifically requested or any information that otherwise qualifies for exclusion.
6 Compare KRS 218A.202, relating to the Kentucky All Schedules Prescription Electronic Reporting System, which provides that "[t]he data and any report obtained therefrom shall not be a public record," and which makes knowing disclosure of the transmitted data to unauthorized persons a Class D felony. KRS 218A.202(8) and (10).
7 See also, "Study Finds 'Huge' Cluster of Lung Cancer in Jefferson County," The Courier-Journal, September 7, 2003 (attached).
8 KRS 214.556(2) provides:
Each licensed health facility which provides diagnostic services, or diagnostic services and treatment, or treatment to cancer patients shall report to the Kentucky Cancer Registry, through the cancer patient data management system and in a format prescribed by the Kentucky Cancer Registry, each case of cancer seen at that health facility. Failure to comply may be cause for assessment of an administrative fine for the health facility . . . .
9 Indeed, a zip code is a much larger geographic universe than an individual school.
10 We have located at least one example of a published report dealing with cancer incidence in a particular zip code. See, "Summary of Cancer Incidence and Mortality for Zip Code 29506 (Florence, SC)." www.scdhec.net/co/phsis/biostatistics/SCCCR/pdfdocs/clusters/29506.pdf. (attached); see also, "National Cancer Database Survey of Breast Cancer Management for Patients from Low Income Zip Codes." jncicancerspectrum.oupjournals.org/cgi/cancerlit/10679664 (attached).