Sue K. RABOVSKY, Appellant,
v.
COMMONWEALTH of Kentucky, Appellee.
No. 96-SC-462-MR.
Supreme Court of Kentucky.
February 19, 1998. Rehearing Denied September 3, 1998.*7 John T. McCall, Louisville, for appellant.
A.B. Chandler, III, Attorney General, and Samuel J. Floyd, Jr., Criminal Appellate Division, Office of Attorney General, Frankfort, for appellee.
COOPER, Justice.
Appellant was convicted of murder in the Jefferson Circuit Court and sentenced to twenty-five years imprisonment. She appeals to this court as a matter of right. Ky. Const. § 110(2) (b).
At 9:40 a.m. on March 18, 1995, Appellant called Jefferson County Emergency Medical Services (EMS) and reported that her husband, Charles Rabovsky, was "not responding." When EMS personnel arrived at the Rabovsky residence, Appellant's husband was unconscious. He was transported to Audubon Hospital in Louisville where he remained comatose until he expired on March 29, 1995. The primary treating physician, Dr. Rukmaiah C. Bhupalam, a neurologist, diagnosed the cause of the victim's comatose state and ultimate death as hypoglycemia due to external administration of a massive dose of insulin. This diagnosis was confirmed by Dr. John Cyrus, a specialist in internal medicine and endocrinology, and by Dr. William Smock, an assistant medical examiner trained in clinical forensics. There was substantial circumstantial evidence that Appellant had the motive, opportunity and animus to inject a lethal dose of insulin into her husband's body, presumably while he was asleep. Appellant raises three claims of error.
I. BLOOD TEST RESULTS
Dr. Bhupalam ordered that blood samples be collected from the victim at periodic intervals. The first sample was collected in the emergency room at 11:22 a.m. on March 18, 1995. Another sample was collected in the intensive care unit at 9:40 p.m. on the same day. Other samples were collected on March 19, March 20, March 23, and March 26, 1995. Audubon Hospital had a contract with National Health Laboratories, Inc. (N.H.L.), located in Louisville, to perform laboratory tests on blood samples collected from patients at the hospital. Although Mr. Rabovsky's blood samples were sent to N.H.L. for testing, the actual tests were performed by National Reference Laboratory (N.R.L.) in Nashville, Tennessee. N.H.L. and N.R.L. are owned by the same holding company, but are different corporate entities. The record does not reveal how N.R.L. reported its test results to N.H.L., but the results were reported to Audubon on N.H.L.'s computer-generated report forms. Although Audubon's medical records were introduced into evidence pursuant to KRE 803 (6), Appellant objected to various specific entries in those records, including N.H.L.'s printouts reporting the blood test results. No evidence was introduced to prove who collected the blood samples, how they were stored, how they were transported to N.H.L., how they were transported (if they were) to N.R.L., or what method was used to test the samples. The samples collected on March 18 were sent to N.H.L. on March 19. Audubon's records contain the following nurses' notes recorded on March 19, 1995:
9:00 a.m./ Dr. Bhupalam called ordered that all existing lab samples be Sent for insulin, C-peptide levels, including original ER lab samples.
9:05 a.m./ lab[1] called informed about MD's orders. RN was informed that results from ER sample may be inaccurate but all samples would be sent. MD called.
9:10 a.m./ Dr. Bhupalam returned call informed about lab stating ER sample may result in inaccurate level but labs would be sent.
The medical experts in this case testified that insulin is produced naturally within the body by the pancreas; and that when it produces insulin, the pancreas also produces a substance called C-peptide. Thus, if the level of naturally produced insulin is abnormally high, the C-peptide level also should be abnormally high. If the insulin level is abnormally high, but the C-peptide level is normal or low, the insulin must have been *8 introduced into the body from an external source. The normal blood insulin level is 5.0-20.0 uU/ML (microunits per milliliter), and the normal level for C-peptide is 1.0-3.0 NG/ML (nanograms per milliliter). The test results of the two blood samples collected on March 18, 1995 were reported on computer printouts dated March 22 and March 23, 1995. Although the reports do not reflect the times when the samples were collected, it was assumed at trial that the March 23 printout reported the results of the sample taken in the emergency room at 11:22 a.m. and the March 22 printout reported the results of the sample taken in the intensive care unit at 9:40 p.m. The printouts each reported the blood insulin level as ">240" (greater than 240 uU/ML). The March 23 printout reported the C-peptide level at "240." On the March 22 printout, someone has handwritten the number "483" next to the computer-generated number ">240." Dr. Bhupalam testified over objection that the handwritten numbers reflected the actual insulin levels in the victim's blood when the first two samples were collected.
Appellant asserts error in admitting the blood test reports containing the handwritten annotations and in permitting Dr. Bhupalam to base his opinion on those reports. However, that error was rendered harmless when the Commonwealth produced Dr. Ronald Wagner, technical director of Laboratory Corporation of America (formerly N.H.L.), who testified without objection[2] that he had checked N.H.L.'s records and determined that the exact blood insulin level on the first test was 1672 microunits per milliliter and that the exact blood insulin level on the second test was 483 microunits per milliliter. Although Dr. Wagner was not the custodian of N.H.L.'s records, he was a "qualified witness" to testify to the contents of those records as permitted by KRE 803(6).
The more serious and ultimately fatal problem with respect to the admission of the blood test results is the total failure of the Commonwealth to establish a chain of custody of the blood samples. This issue relates to the integrity of the evidence and is an integral part of the authentication requirement of KRE 901 (a). The purpose of requiring proof of the chain of custody of a blood sample is to show that the blood tested in the laboratory was the same blood drawn from the victim. R. Lawson, The Kentucky Evidence Law Handbook, § 11.00, p. 592 (3rd ed. Michie 1993). While the integrity of weapons or similar items of physical evidence, which are clearly identifiable and distinguishable, does not require proof of a chain of custody, e.g., Beason v. Commonwealth, Ky., 548 S.W.2d 835 (1977), Smith v. Commonwealth, Ky., 366 S.W.2d 902 (1962), a chain of custody is required for blood samples or other specimens taken from a human body for the purpose of analysis. Henderson v. Commonwealth, Ky., 507 S.W.2d 454 (1974); Calvert v. Commonwealth, Ky. App., 708 S.W.2d 121, 124 (1986); Haste v. Kentucky Unemployment Ins. Comm'n, Ky. App., 673 S.W.2d 740 (1984); Lawson, supra, § 11.00, p. 593; 32A C.J.S. Evidence § 797 (1996).
Even with respect to substances which are not clearly identifiable or distinguishable, it is unnecessary to establish a perfect chain of custody or to eliminate all possibility of tampering or misidentification, so long as there is persuasive evidence that "the reasonable probability is that the evidence has not been altered in any material respect." United States v. Cardenas, 864 F.2d 1528, 1532 (10th Cir. 1989), cert. denied, 491 U.S. 909, 109 S. Ct. 3197, 105 L. Ed. 2d 705 (1989). See also Brown v. Commonwealth, Ky., 449 S.W.2d 738, 740 (1969). Gaps in the chain normally go to the weight of the evidence rather than to its admissibility. United States v. Lott, 854 F.2d 244, 250 (7th Cir. 1988). Here, however, there was no attempt at all to establish the chain of custody *9 of these blood samples, even though the samples apparently were transferred and stored internally within the hospital, then transferred and stored outside the hospital, first at a laboratory in Louisville, then, presumably, at another laboratory in Nashville. As Justice Palmore aptly put it in Henderson v. Commonwealth, supra:
Hence the integrity of the evidence from the time it was relinquished by the investigative officers until it reached the laboratory analyst was not proved. We think that surely it is unnecessary to delve into the literature of the law in order to document the point that this type of carelessness in the development of important evidence during the course of a trial simply will not do. We know it is tedious and time-consuming to trace the integrity of an exhibit; in fact, it is tedious and time-consuming to have a trial at all when we think we know the defendant is guilty anyway, but it is not half as bad a nuisance to do it right the first time as it is to go through the whole process a second time two years later.
Id. at 461.
The dissenting opinion's reliance on Buckler v. Commonwealth, Ky., 541 S.W.2d 935 (1976) is misplaced. With respect to the admission of hospital records, Buckler reiterated the holdings in Bellew v. Commonwealth, Ky., 477 S.W.2d 779 (1972) and Whittaker v. Thornberry, 306 Ky. 830, 209 S.W.2d 498 (1948) that such records were admissible under the common law "shopbook" exception to the hearsay rule. The records in Buckler were written reports prepared by doctors who had treated the defendant's mental illness during her stay at Central State Hospital. The trial court excluded the records because the doctors who originally prepared the reports had not been subpoenaed; thus, there was no proof of "necessity" to admit the reports through the shopbook exception, as had been required by Whittaker v. Thornberry, supra, 209 S.W.2d at 501. In reversing, we overruled the "necessity" aspect of Whittaker, i.e., that application of the shopbook exception required proof of the actual unavailability of those who prepared the entries in question.
The only records requiring authentication in Buckler were the written records of treatment rendered within the hospital by the doctors who authored the records. Proof of entry at or near the time of occurrence by one with knowledge and during the regular course of business suffices to authenticate this type of documentary evidence. KRE 902(11)(A). Buckler did not address the need to further authenticate real evidence, such as blood samples, because that issue was not presented by the evidence in that case. Nor was it necessary in Buckler to address the legal principle that an entry in a business record is not admissible simply because it is in a business record, if that entry would be inadmissible for another reason. Prater v. Cabinet for Human Resources, Ky., 954 S.W.2d 954, 958 (1997); Alexander v. Commonwealth, Ky., 862 S.W.2d 856, 862 (1993). Although KRE 803 (6), the successor to the common law shopbook exception, is available to admit a report of a blood test result without the necessity of producing the person who prepared the report, the report is inadmissible if it could not have been introduced by the person who prepared it. Here, the persons who prepared these blood test reports could not have testified to their contents absent a preliminary showing sufficient to satisfy the authenticity requirement of KRE 901 (a) that the blood which was tested was that of the patient in question. That preliminary showing would have required proof of the chain of custody of the samples tested. The fact that the blood test reports were ultimately placed in the business records of Audubon Hospital does not alter this requirement.
The quote from Brown v. Commonwealth, supra, set forth in the dissenting opinion herein, is taken out of context. Immediately preceding that quote is the following:
The evidence established that a deputy coroner took the blood sample, labeled it, placed it in a manila envelope, and delivered it to Coroner Hager, who in turn delivered it to "Mr. Cash" at the University of Kentucky Medical Center. Mr. Cash, under the direction of Dr. Wilmer M. Talbert, Jr., pathologist, made the analysis.
*10 Id. at 740. If similar evidence had been introduced in the trial of this case, there would be no issue with respect to the integrity of the blood samples tested at N.R.L.
Since the results of the analyses of the blood samples were introduced without establishing the integrity of the samples by showing the chain of custody, the judgment of conviction must be reversed for a new trial.
II. NURSES' NOTES
Because he suspected Appellant was the culprit who had injected the lethal dose of insulin into her husband's body, Dr. Bhupalam instructed the hospital nurses to record in the nurses' notes anything they overheard Appellant say about her husband. The nurses dutifully recorded that they overheard Appellant state that she was "through mourning," that she had done her mourning "last Thursday," that her husband was not "here," that he was "in heaven," and that his body was "the only thing left." (He was still alive when these remarks were made.) The nurses also recorded that Appellant told them that her husband had been out drinking with friends on the evening before his admission to the hospital, a statement which conflicted with evidence that the victim had been on National Guard duty that night. All of these notes were admitted into evidence as "business records" of the hospital. KRE 803(6).
To reiterate what was said with respect to the blood test results, an individual entry in a business record is not admissible simply because it is included in a business record, if the entry would be inadmissible for another reason. Thus, if Appellant's statements would have been inadmissible if the nurses had testified at trial, they were likewise inadmissible when offered as entries in the hospital records. There was a great debate at trial as to whether the underlying statements were admissible under KRE 803(4) as statements made for the purpose of treatment or diagnosis. They were not. KRE 803(4) pertains only to statements "reasonably pertinent to the treatment or diagnosis" of the declarant or an infant child or ward of the declarant. Prater v. Cabinet for Human Resources, supra, at 959. Obviously, these statements were not "reasonably pertinent to the treatment or diagnosis" of Appellant, the declarant. However, Appellant is a party to this action, and any statement made by a party is admissible against that party as an admission. KRE 801A(b)(1). If the nurses had testified at trial, there would have been no issue as to the admissibility of these statements.
However, the nurses did not testify, and the statements were not admissible as "business records" of the hospital. KRE 803(6) requires inter alia that (1) both the maker of the record and the person providing the information for the record must have been acting under a business duty to do so, Prater v. Cabinet for Human Resources, supra, at 959, Alexander v. Commonwealth, supra at 861, Lawson, supra, § 8.65 V, pp. 465-66; and (2) it must have been the regular practice of the business in question to make the memorandum, report or record. Martin v. Taylor's Ex'x, 285 Ky. 128, 147 S.W.2d 70, 72 (1941); Lawson, supra, § 8.65 II, pp. 461-62.
With respect to the "business duty" requirement, some of Appellant's statements, e.g., the victim's activities on the evening prior to his hospitalization, were intended to provide information which Appellant was not under a business duty to provide. Those statements were clearly inadmissible as business records. Other statements, e.g., that Appellant was "through mourning," were not intended to provide information, but were nonhearsay statements tending to prove the state-of-mind of the declarant. Lawson, supra, § 8.05, pp. 364-65. The nurses who recorded the statements were providing information from their observation and hearing and, ostensibly, were under a business duty to report the information, because they had been directed to do so by Dr. Bhupalam.
However, none of the statements satisfy the "regular practice" requirement. "The reliability of business records has always been thought to depend heavily on their use in the regular course of conducting business." Lawson, supra, § 8.65 II, pp. 461-62. The abiding principle is that the entry offered must be a part of a series of entries or reports, not a casual or isolated one. Hiram *11 Ricker & Sons v. Students Int'l Meditation Soc'y, 501 F.2d 550, 554 (1st Cir. 1974). With respect to medical records, this exception is premised on the admissibility of information "important to an effective diagnosis or treatment." Souder v. Commonwealth, Ky., 719 S.W.2d 730, 735 (1986). We held in Souder that this does not include information provided as part of a criminal investigation. The nurses' notes in question in this case were not important to an effective diagnosis or treatment of the patient, but were recorded only for the purpose of attempting to prove that Appellant killed her husband. It was not the regular practice of the nurses to make such entries; thus, the entries were casual or isolated and not within the regular course of business of the hospital. Although the nurses could have testified in person to the statements which they overheard, the statements were not admissible through the medium of KRE 803 (6).
III. TESTIMONY OF DR. WILLIAM SMOCK
Dr. Smock is not on the staff of Audubon Hospital, but is an assistant medical examiner trained in clinical forensics. He personally examined the victim prior to his death and reviewed the medical records and laboratory reports. He concurred in Dr. Bhupalam's diagnosis of hypoglycemia due to introduction of insulin from an outside source. He also read from selected entries in the medical records and Appellant claims that such violated the proscription against hearsay. However, the medical records were properly introduced under KRE 803(6). If the entries read by Dr. Smock were not otherwise inadmissible, Appellant was no more prejudiced by the fact that the doctor read from the records than she was by the fact that the records were introduced in the first place.
Even if any specific entries in the records were inadmissible, Dr. Smock could have "reasonably relied" on those entries in forming his expert opinion as to the diagnosis and cause of the victim's condition. KRE 703(a). If so, it would have been within the discretion of the trial judge to permit disclosure of the information to the jury (accompanied by an appropriate admonition, if requested), assuming the judge determined that the information was "trustworthy, necessary to illuminate testimony, and unprivileged." KRE 703 (b). The applicability of KRE 703 (a) and (b) to this evidence was not addressed at the first trial, because the trial judge deemed the hospital records to be admissible in toto. It was error to permit Dr. Smock to read from inadmissible entries without addressing the factual determinations required by KRE 703 (b). See KRE 104 (a). This error should not recur upon retrial.
For the reasons stated in this opinion, the judgment of conviction and the sentence imposed upon Appellant are reversed and this action is remanded to the Jefferson Circuit Court for a new trial.
STEPHENS, C.J., and JOHNSTONE and STUMBO, JJ., concur.
GRAVES, J., dissents by separate opinion, with LAMBERT and WINTERSHEIMER, JJ., joining that dissent.
GRAVES, Justice, dissenting.
Respectfully, I must dissent. Under the circumstances of this case, the blood test results should be admissible as reliable evidence. The blood samples were initially taken for medical diagnosis and treatment and not for a criminal prosecution. Had criminal prosecution, such as DUI (where blood is drawn at the hospital and then sent to a state police laboratory), been the primary purpose for taking and analyzing the blood samples, I would concur with Justice Cooper. However, these blood samples were routinely taken and analyzed in accordance with modern everyday hospital practices, presumably for the primary purpose of saving life. There is a presumption that the vial of blood carried the decedent's identification from the hospital to the laboratory. To require proof of chain of custody of blood samples or other body fluids should not be necessary when these same results are used by medical professionals for the purposes of making life and death determinations in arriving at decisions concerning diagnosis and treatment. It is not reasonable to hold a treating physician, operating under exigent circumstances to save a life, to *12 the same standards as a trained detective in collecting and preserving evidence for criminal prosecution.
[T]he calling of all the individual attendant physicians and nurses who have cooperated to make the record even of a single patient would be a serious interference with convenience of hospital management. There is a Circumstantial Guarantee of Trustworthiness; for the records are made and relied upon in affairs of life and death. Moreover, amidst the day-to-day details of scores of hospital cases, the physicians and nurses can ordinarily recall from actual memory few or none of the specific data entered; they themselves rely upon the record of their own action; hence, to call them to the stand would ordinarily add little or nothing to the information furnished by the record alone. . . .
Buckler v. Commonwealth, Ky., 541 S.W.2d 935, 938 (1976).
Further, requiring proof of a chain of custody before allowing the admissibility of the results of analysis of body fluids by a hospital laboratory would nullify the benefits of KRS 422.305 and KRS 422.310. Moreover, hospitals may likely be reluctant to assist or cooperate in investigations of crimes due to time and cost. In this case, satisfying the chain of custody requirements may have taken additional time when time was of the essence for commencing treatment.
The proof of authenticity may come in any form. Here, the medical autopsy proved that the decedent died from hypoglycemia due to an insulin overdose. Since the blood analyses proved accurate in the diagnosis, which was confirmed by the response to the subsequent treatment and the ultimate demise of the decedent, these results should constitute a prima facie showing of authenticity of the blood samples with the burden of proof being on the challenger to show lack of authenticity of same.
The fact that an autopsy confirmed the diagnosis of death due to hypoglycemia resulting from insulin overdose authenticates the hearsay data concerning the blood samples. Therefore, this data should be legally sufficient proof of the reliability of the blood samples even though chain of custody was not established. A hand written notation of 1672 micro units per milliliter should not affect the admissibility, because the printout showed an amount greater than 240 micro units per milliliter when anything over 20 micro units per milliliter indicated a pathological condition consistent with the final proven diagnosis.
In Buckler, supra, this Court established the following standard by which to measure the admissibility of opinion testimony that is based, in part, on hearsay evidence.
[W]e adopt as an exception to the hearsay rule in Kentucky that an expert may properly express an opinion based upon information supplied by third parties which is not in evidence, but upon which the expert customarily relies in the practice of his profession. Our previous opinions to the contrary are hereby overruled. We emphasize that the type of information which can be utilized by the expert in forming his opinion would be only that produced by qualified personnel and on which the expert would customarily rely on in the day-to-day decisions attendant to his profession. Such a limitation, we feel, guarantees a relatively high degree of reliability and frees the expert to use for his testimony the tools on which he normally relies in making a diagnosis.
Judge Palmore's comments in Henderson v. Commonwealth, Ky., 507 S.W.2d 454 (1974), should not apply in this case because we have impersonal, high volume, mass produced managed health care, not careless investigatory procedures by law enforcement officers. As our highest Court stated in Brown v. Commonwealth, Ky., 449 S.W.2d 738, 740 (1969), "We cannot visualize any motive on the part of those who handled the blood sample to tamper with it. That there was any tampering is the barest speculation, which is not enough to destroy its integrity." Here, ample indicia of reliability make it improbable that the blood analyzed was from anyone other than the decedent.
In my view this is a needless reversal for insubstantial reasons, contrary to settled Kentucky law, which introduces a non-substantive *13 procedural barrier into the criminal justice system. There is no need to require more precise identification of blood when it is used as evidence in a criminal case than the clinician requires when using the results of blood testing for diagnosis and treatment. In circumstances such as these, we should have sufficient confidence in the integrity and proficiency of the medial profession to require no greater standard than it requires of itself.
LAMBERT, and WINTERSHEIMER, JJ., join this dissenting opinion.
NOTES
[1] Presumably the "lab" referred to in these notes was the hospital's own laboratory, since the notes were entered before the samples were sent to N.H.L.
[2] We emphasize "without objection," because it is unclear from the record why N.H.L.'s records would be relevant to prove the results of tests performed at N.R.L.