Theofanis v. Sarrafi

791 N.E.2d 38, 339 Ill. App. 3d 460, 274 Ill. Dec. 242

Opinion

JUSTICE McNULTY

delivered the opinion of the court:

Ghodratollah Sarrafi, M.D., participated in a health maintenance organization (HMO) of Health Care Service Corporation (HCSC). On June 3, 1996, Sarrafi received the results of a test that showed a dangerous mass in the heart of Sofia Beniamin, one of Sarrafi’s patients. Eight days later Sofia suffered a stroke that left her unable to speak. Sofia’s daughter, as her guardian, and Sofia’s husband sued Sarrafi and HCSC for failing to inform Sofia of the results of the test.

At the jury trial the court allowed Sarrafi to testify from notes he made of conversations with Sofia. The jury found both defendants liable but assessed $0 in damages. The trial court amended the verdict to a finding in favor of Sarrafi. The court also entered judgment notwithstanding the verdict in favor of HCSC because the court found the evidence of agency insufficient.

We hold that the inconsistent verdict, which may have resulted from compromise, requires remand for retrial. The Dead-Man’s Act (the Act) (735 ILCS 5/8—201 (West 2000)) precludes the testimony concerning conversations with Sofia, even if the doctor made contemporaneous notes of the conversations. The plaintiffs presented sufficient evidence to present to a jury on the issue of implied agency of Sarrafi for HCSC. Therefore, we reverse the judgment of the trial court and remand for retrial.

BACKGROUND

Mooshi and Sofia Beniamin immigrated to the United States from Iran, with their three daughters, in 1979. Mooshi found work at a manufacturing plant and obtained health insurance for his family through that job. Sofia fell ill in 1985. The insurance paid for her visit to Sarrafi, who had also immigrated to the United States from Iran. Sarrafi noted the stenosis, or narrowing, of a valve in Sofia’s heart. She continued to see Sarrafi fairly regularly over the following years. In 1986 Sarrafi diagnosed Sofia’s hypertension, and in 1991 he also discovered that she suffered from atrial fibrillation, which is an irregular heartbeat.

In 1995 Mooshi’s employer changed the insurance options available to its employees. The new insurer, HCSC, allowed employees three options. They could receive traditional insurance, or they could enroll in the BlueAdvantage HMO or their preferred provider organization (PPO). Mooshi chose the HMO, and he chose Sarrafi, from HCSC’s fist, as the family’s primary care physician. HCSC had a contract with Holy Family Physicians Organization (Holy Family), and Sarrafi belonged to that organization. HCSC compensated Holy Family for the health care it provided, and Holy Family paid Sarrafi for his work.

At an office visit in 1996, Sarrafi arranged an echocardiogram (EKG) for Sofia. Dr. Vupparahalli Ramesh performed the EKG on June 3, 1996. When he received the films from the test later that day, Ramesh discovered “a mass effect near the apex [of the heart] ***, consistent with the presence of thrombus.” A thrombus is a blood clot. Small particles, or emboli, can break off a thrombus in the heart at any time. The emboli from the thrombus can then course through the bloodstream and lodge in distant parts of the body, cutting off the flow of blood to those areas. Emboli in the brain may cause a stroke.

Ramesh promptly called Sarrafi to report the finding. Ramesh suggested that Sofia should undergo a transesophageal EKG to determine in better detail the nature of the mass. Sarrafi did not tell Sofia about the EKG results or the need for a transesophageal EKG. Sofia called Sarrafi’s office on June 8, 1996, but she did not hear the results of the EKG or of the need for further tests. Sarrafi did not speak to Sofia between June 3, 1996, and June 11, 1996.

Sofia suffered a severe stroke on June 11, 1996. The stroke left her unable to speak or walk. Although Sofia eventually regained very limited ability to walk, she never regained the power of speech, not even to the extent of nodding to mean yes in response to questions. A court found Sofia legally disabled and appointed Sofia’s eldest daughter, Glorinda Theofanis, plenary guardian of Sofia’s estate and person.

In December 1996 Glorinda, in her role as guardian of Sofia’s estate, brought this lawsuit. After several amendments to the complaint, Glorinda and Mooshi reduced their claims to four counts. In the first count Glorinda alleged that Sarrafi negligently failed to inform Sofia of the results of the EKG, negligently failed to order further diagnostic tests, and negligently failed to hospitalize Sofia and treat her with Heparin, a blood-thinning medication that could dissolve the clot, after he learned the results of the EKG. Glorinda sought to recover damages from HCSC, as well as Sarrafi, on the theory that Sarrafi acted as an apparent or implied agent of HCSC. In the second count Mooshi pled the same facts and sought recovery from the same parties for his loss of consortium. In count III Glorinda alleged that the contract between HCSC and Holy Family created incentives for physicians affiliated with Holy Family to minimize the hospitalization of patients and the number of diagnostic tests performed. She alleged that Sarrafi breached his fiduciary duty to disclose the financial incentives. In count IV Mooshi alleged that the same breach of fiduciary duty led to his loss of consortium.

Sarrafi denied the essential charges of the complaint and added an affirmative defense that Sofia negligently contributed to her injury by refusing the reasonable advice of her physician. Sarrafi alleged that at an office visit in May 1996 Sofia declined Sarrafi’s offer to hospitalize her and administer Coumadin, a blood-thinning medication related to Heparin. In support of the allegations, Sarrafi attached to the answer handwritten notes Sarrafi made concerning the office visits in May 1996.

At trial plaintiffs called Sarrafi as an adverse witness. Sarrafi admitted that Ramesh called him on June 3, 1996, following the EKG, and told Sarrafi that he found a mass in Sofia’s heart that could be a tumor or a thrombus. Sarrafi also admitted that in the written report he received shortly thereafter, Ramesh said he found a “soft ***. friable mass, giving the appearance of a thrombus.” Either a tumor or a thrombus could embolize at any time, but a thrombus carries a greater risk of embolization.

According to Sarrafi, he tried to call Sofia to tell her about the test results immediately after he heard from Ramesh. Either the call went unanswered or the line was busy. Over the following days Sarrafi tried to call Sofia five or six times, but he never got through. He did not ask anyone on his staff to call Sofia. Although all of Sofia’s daughters were Sarrafi’s patients, he never tried to contact any of them. He sent no letter concerning the results of the EKG.

Sarrafi’s attorney asked Sarrafi on cross-examination to read to the jury the notes Sarrafi made after Sofia came to his office on May 28, 1996. Plaintiffs objected that the testimony would violate the Act. The court overruled the objection. Sarrafi testified that in his notes he wrote:

“Patient refused hospitalization and Coumadin ***. The importance of taking Coumadin due to atrial fibrillation was discussed with the patient in the presence of Herminda [Spencer, Sofia’s] daughter. She still refused to take it.”

Sarrafi testified that, in his opinion, Sofia faced about an 80% probability of having a stroke within a year if she had no medication.

Glorinda testified that as of June 3, 1996, her parents had call waiting and a functioning answering machine.

Plaintiffs’ expert, Dr. Franklin Wefald, testified that the applicable standard of care required Sarrafi to notify Sofia immediately about the EKG results and to warn her that she faced a potentially devastating complication if she did not start taking a strong blood thinner. In Dr. Wefald’s opinion, Sarrafi violated the standard of care by failing to use all available means to contact Sofia. Given the severity of the situation, Sarrafi should have asked police to go to Sofia’s home if he tried to call her and received no answer. If Sarrafi had contacted Sofia he should have emphasized the severity of the risk of stroke and the necessity of hospitalization and the use of anticoagulants like Heparin. According to Wefald, if Sarrafi had given Sofia appropriate advice based on the EKG, the anticoagulants likely would have prevented Sofia from suffering the stroke.

On cross-examination Wefald estimated about a 20% to 25% probability that a patient with atrial fibrillation and mitral valve stenosis would suffer a stroke within a year without anticoagulant therapy. He estimated that the probability of stroke doubled or tripled in the presence of a thrombus in the heart. Thus, his estimate of the risk Sofia faced nearly matched Sarrafi’s estimate of that risk. Defense counsel suggested that the risk of a stroke in the week from June 3 to June 11, 1996, was therefore about 1%, which he arrived at by dividing 2 x 25% by 52 weeks in a year. Wefald answered that the simple division of the probability did not accurately represent the probability for the week.

Plaintiffs asked Wefald to respond to Sarrafi’s testimony about his handwritten notes from May 28, 1996. Wefald said that the notes “were medically inconsistent within the structure of the entire record for that day.” He explained that the notes in the body of the page referred to Sofia’s complaints and Sarrafi’s assessment of the atrial fibrillation. Then, on the margins of the page, Sarrafi added a note charging Sofia with refusing hospitalization and Coumadin. The marginal comment did not make sense in light of Sofia’s actual complaints and prior course of treatment, which would not have suggested an immediate need for hospitalization or Coumadin. To Wefald, the marginal notes appeared to have been written later, probably after Sofia suffered the stroke.

Plaintiffs’ second expert agreed that, even assuming the truth of Sarrafi’s testimony, the standard of care required greater efforts to contact Sofia. The expert also opined that if Sarrafi had contacted Sofia, the further testing, if not the EKG alone, would have persuaded Sofia to accept Heparin and Coumadin treatment. The medication would have significantly reduced the risk of stroke, at least cutting the risk by a factor of two. The expert testified that some in his profession would say the blood thinners could reduce the risk of stroke by 80%.

One of defendants’ experts agreed that a blood thinner like Coumadin would “decrease the chance of stroke by over half, maybe two-thirds.” He accepted the defense attorney’s suggestion that the probability of a stroke occurring within the week between June 3 and June 11, 1996, was 1/b2 times the probability of a stroke within a year. The expert believed Sarrafi met the standard of care by trying to call Sofia five or six times between June 3 and June 11, 1996.

Another defense expert testified that he read the original EKG films, and he disagreed with P.amesh’s report. The mass shown on the EKG was probably a tumor, not a thrombus. While the tumor might embolize and cause a stroke, neither Coumadin nor Heparin would affect the risk of such a stroke.

On the issue of agency, plaintiffs presented testimony from several current and former employees of HCSC. The first witness admitted that an HCSC committee determined whether each doctor seeking to join its network met HCSC’s criteria for inclusion. HCSC had the right to exclude any doctor from its network, even if HCSC had a contract with a medical group that included the doctor.

HCSC also evaluated every doctor in its network every two years. In the evaluation an HCSC employee would go to the doctor’s office and pull a random sample of his medical charts. If HCSC found certain problems with the medical charts, it could remove the doctor from its network. The HCSC employee would look at the quality of care the doctor provided, along with other factors.

But HCSC had no prior control over a participating doctor’s treatment decisions, including decisions to refer a patient to a specialist, to perform medical tests, or to hospitalize the patient. Sarrafi did not need HCSC’s prior approval for any treatment he ordered. HCSC paid Holy Family, according to their contract, for all services Holy Family’s member physicians provided to HCSC members. The contract between Holy Family and Sarrafi specified that Sarrafi acted as an independent contractor in the Holy Family organization. Similarly, HCSC’s contract with Holy Family specified that neither acted as an agent of the other.

HCSC’s employees identified several advertising pamphlets HCSC distributed to businesses in an effort to sell them the BlueAdvantage HMO. In the pamphlets HCSC said that each physician in its network met or exceeded their “ quality [,] effectiveness [and] patient satisfaction standards.”

Mooshi testified that he followed his daughter Herminda’s advice about insurance. Herminda testified that she reviewed all the literature concerning insurance choices for her father. When Mooshi’s employer changed its insurer to HCSC, Herminda looked at materials from BlueAdvantage HMO amongst other documents. She advised her father to sign up for the HMO based on a combination of factors, including the representations HCSC made about the quality of its doctors. She would have suggested switching doctors if better insurance coverage restricted choices to other doctors.

In 1990 and 1991, when Mooshi’s employer had other insurance, Sofia received treatment from a doctor in that insurer’s network. But Sarrafi’s records show that Sofia continued coming to Sarrafi’s office for treatment in 1990 and 1991. Sarrafi testified that at least once Sofia, rather than the insurer, paid the $25 fee for the office visit.

Plaintiffs also presented evidence concerning HCSC’s contract and the financial incentives for Holy Family and for Sarrafi as a member of Holy Family, to minimize the number of tests and hospitalizations of its patients. Sarrafi admitted that he did not inform Sofia about the incentives, because he would not allow the incentives to influence his treatment decisions. All of the experts agreed that they would not inform their patients of similar incentive clauses, and they would not allow such incentives to influence the treatments they provided any of their patients.

The parties prepared multiple verdict forms for each count. The court instructed the jurors to return verdict form B for count I only if they unanimously found against Sarrafi and that Sofia was not contributorily negligent and that Sarrafi acted as an agent of HCSC. The court told the jurors that to return a verdict against Sarrafi, the jurors must find that plaintiffs met their burden of proving (1) that Sarrafi acted negligently, (2) that Sofia suffered injury, and (3) “that the negligence of the defendants was a proximate cause of the injury to the plaintiffs.” The court reiterated the requirement: “If you find from your consideration of all the evidence that any of these propositions has not been proved, then your verdict should be for the defendants.” The court instructed the jurors similarly on the other three counts.

After substantial deliberations the jurors sent the court a note, which said:

“It is the general con[s]ensus among this jury that no compromise can be reached. We request that this jury be either re-instructed or dismissed, as no resolution is in sight, and the tenor has become rather argumentative.”

Later the jury foreperson sent a note asking:

“Do we have an obligation to seek the truth — or only consider the evidence even if the truth is only around the corner?”

The court instructed the jurors to continue deliberations.

Shortly thereafter the jury returned verdicts in favor of defendants on counts II, III, and IV of the complaint. That is, the jurors found that plaintiffs did not meet their burden of proof on both of the counts charging Sarrafi with violating a fiduciary duty to disclose financial incentives, and that Mooshi failed to meet his burden of proof on the count seeking damages for loss of consortium caused by Sarrafi’s negligence. On count I the jurors all signed verdict form B, finding in favor of Glorinda and against both defendants, but the jury assessed damages of $0.

After the court entered judgment on the verdict, all parties filed posttrial motions. The court denied plaintiffs’ motion for new trial and granted HCSC judgment notwithstanding the verdict on count I. The court concluded from the award of zero damages that the jury found that Sarrafi’s acts did not proximately cause injury, and on that basis the court amended the verdict to a finding in favor of Sarrafi on count I.

DISCUSSION

I

A

Plaintiffs argue on appeal first that the trial court erred by denying their motion for a new trial. Sarrafi counters that this court should affirm the trial court’s judgment in his favor because the evidence at trial so overwhelmingly supports the judgment that no verdict in favor of the plaintiffs could ever stand because plaintiffs failed to prove his negligence caused the harm. See Pedrick v. Peoria & Eastern R.R. Co., 37 Ill. 2d 494, 502 (1967). We disagree. Plaintiffs presented sufficient evidence supporting their case. Sarrafi’s argument relies on statistical assumptions without support in the record and fallacious statistical reasoning. Even apart from these errors, the evidence on which Sarrafi relies has no bearing on the issue of whether his negligence in fact caused the harm.

Plaintiffs presented evidence that Sarrafi failed to contact Sofia between June 3, 1996, and June 11, 1996. Glorinda testified that Sofia had a functioning answering machine and call waiting. From this evidence the jury could conclude that Sarrafi did not even attempt to call Sofia after he learned the results of the EKG. The experts agreed that after Sarrafi received the call from Ramesh about the EKG, the standard of care required significant efforts to contact Sofia. Plaintiffs’ experts testified that the failure to contact Sofia caused her not to have further tests and caused her not to learn the new evidence of a significant risk of stroke. The jury, like plaintiffs’ experts, could rely on Ramesh’s finding that the mass in Sofia’s heart appeared to be a thrombus, and therefore that mass presented a considerable, and reducible, risk of embolization and stroke. Thus, the failure to contact Sofia for further tests and treatment with Heparin and Coumadin significantly increased the risk of a stroke.

Sarrafi uses the distorted statistical arguments he made at trial as the grounds for his claim that the trial court should have granted him judgment notwithstanding the verdict. At trial Sarrafi’s attorney suggested that if the probability of a stroke within a year is 52%, then the probability of a stroke within a week is 1%. Sarrafi argues that the 1% probability is too small for his acts to count as the cause of the stroke, even if he did nothing to reduce the 1% probability.

The parties could have avoided some of the distortion inherent in this argument if either had consulted any authority on statistics. A court of appeal may take judicial notice of laws of mathematics and computational methods scientists generally accept as irrefutable. Thomas v. Price, 81 Ill. App. 3d 542, 545 (1980); Cook County Department of Environmental Control v. Tomar Industries, 29 Ill. App. 3d 751, 754 (1975); Mueth v. Jaska, 302 Ill. App. 289, 294-95 (1939). Here we take judicial notice of the proper methods for computation of probabilities of a conjunction of events given the probabilities of each separate event.

The probability of two independent events A and B is the product of the probabilities (P(A) x P(B)), not the sum of the probabilities, as Sarrafi’s method presumed. The probability of neither of two independent events is (1-P(A)) x (l-P(B)). If the probability of a stroke in each week is 1%, and the probability each week is independent of the probability all other weeks, the probability of a stroke within two weeks is not 1% + 1%; instead it is 1 - (,99)x(.99), which is one minus the product of the probability of no stroke in the first week times the probability of no stroke in the second week. If the chance of having a stroke within a year without medication is 80%, as Sarrafi estimated, and the risk is uniformly distributed throughout the year, and the risk each week is independent of the risk for each other week (no expert testimony supported these implicit assumptions), the risk of having a stroke within one week is not .8 divided by 52; instead, it is 1- the 52nd root of .2, which calculates to approximately a 3% risk each week ((l-.03)52=(l-.8)). For a fuller explanation see 1 W Feller, An Introduction to Probability Theory and Its Applications, at 47-49 (3d ed. 1968), or any other elementary treatise on probability and statistics.

Thus, if defendants had presented admissible evidence of statistical independence of each week’s risk of stroke, and if defendants had presented admissible evidence that the risk each week matched the risk every other week, then the evidence would support an estimate of a 3% chance of a stroke during the week at issue. Because the evidence at trial lacked any support for the implicit assumptions about uniformity and independence, it cannot support such a low estimate of the risk. The evidence actually presented, even if augmented with support for the implicit assumptions, does not in any way support a finding of a 1% risk of stroke for the week.

If plaintiffs had objected to Sarrafi’s questions based on the unsupported assumptions and fallacious reasoning, the trial court should have sustained the objections. See Foss Park District v. First National Bank of Waukegan, 125 Ill. App. 2d 276, 280 (1970). The distorted statistical argument illustrates the warning that “[a] lawyer who acts as his own statistician is as worthwhile to the case and the client as a statistician who acts as the trial lawyer.” J. Kobayashi, Killing Them Softly with Your Song: Problems with Proof of Causal Relationship by Statistical Methods and Probability Theory and Expert Opinions with Suggested Methods for Analysis and Cross-examination, and Notes about Epidemiological Studies and Animal Data as Causation Evidence, 363 PLI/LIT 37, 70 (1988).

Even without the fallacious reasoning about probabilities, the evidence had no bearing on the issue of causation in fact. If a negligent act creates only 1 chance in 100 of harm, the act in fact causes the harm in that 1 case in 100 when the chance is realized and the harm occurs. The foreseeability of harm — the ex ante risk — is an important factor for determining whether the defendant had a duty to the plaintiff and whether to impose liability on the defendant for breach of duty. See Lee v. Chicago Transit Authority, 152 Ill. 2d 432, 455-56 (1992). Expert testimony in this case established that the ex ante risk here rose to a level that imposed on Sarrafi a duty to contact Sofia. Sofia’s condition, including the mass found in her heart, made a stroke readily foreseeable in the absence of medication.

To determine cause in fact, courts usually look to the precise sequence of events and such issues as whether the harm would likely have occurred in the absence of the negligent acts. See W. Keeton, Prosser & Keeton on Torts § 41, at 266 (5th ed. 1984). Where a defendant’s acts or omissions have increased the risk of a harm that later occurs, courts look to the relative risk — the ratio of the risk with the negligent act to the risk without negligence — to decide whether the negligent acts or omissions constitute a cause in fact of the harm. Several courts have concluded that a plaintiff meets the burden of proving causation by presenting evidence that the relative risk due to the defendant’s acts or omissions exceeds two. E.g., Allison v. McGhan Medical Corp., 184 F.3d 1300, 1315 n.16 (11th Cir. 1999); see L. Finley, Guarding the Gate to the Courthouse: How Trial Judges Are Using Their Evidentiary Screening Role to Remake Tort Causation Rules, 49 DePaul L. Rev. 335, 348-53 (1999) (arguing that courts should not exclude as irrelevant evidence of a relative risk between one and two, because other evidence may show that the increased risk formed a substantial factor in bringing about the actual harm suffered). Where the risk with the negligent act is at least twice as great as the risk in the absence of negligence, the evidence supports a finding that, more likely than not, the negligence in fact caused the harm.

Here, plaintiffs’ expert testified to a relative risk between two and three. Defendant’s expert agreed, as he said that proper medication would have reduced Sofia’s risk by half to two-thirds. Plaintiffs’ other expert testified that some would put the relative risk as high as five, as he said that the medication might reduce the risk by 80%. Thus, plaintiffs presented ample evidence from which a jury could conclude that Sarrafi’s negligent failure to call, and the consequent failure to further test and medicate Sofia appropriately in light of the known risk, caused the particular harm she suffered. The record cannot support entry of judgment notwithstanding the verdict in favor of Sarrafi.

B

Plaintiffs contend that the trial court should have granted them a new trial on damages only because the jury awarded plainly inadequate damages. We will reverse the trial court’s denial of a motion for new trial only if the trial court abused its discretion. Hulbert v. York, 319 Ill. App. 3d 54, 57 (2001). The court should order a new trial limited to damages only if the evidence amply supports the verdict on liability, the issue of liability is sufficiently separate from the amount of damages, and the record does not suggest a compromise verdict. Hollis v. R. Latoria Construction, Inc., 108 Ill. 2d 401, 408 (1985).

As defendants point out, the record here strongly suggests a compromise verdict. The note jurors sent to the court directly informed the court that the jurors had tried to reach a compromise. The court, by instructing the jurors to continue deliberations, did not attempt to dissuade them from reaching a compromise. By the verdict the jury formally found that Sarrafi acted negligently and his negligence caused plaintiffs damages, but the jury also awarded $0 damages. The verdict itself suggests a compromise. See Taylor v. Manhattan Township Park District, 138 Ill. App. 3d 23, 27 (1985). The trial court did not abuse its discretion by denying the motion for a new trial on damages only.

But the trial court erred by changing the verdict from one in favor of Glorinda on the issue of liability to a finding in favor of Sarrafi. The trial court has discretion to amend the jury’s verdict (Smith v. City of Evanston, 260 Ill. App. 3d 925, 937-38 (1994)), but the court abuses that discretion if it ignores recognized legal principles (State Farm Fire & Casualty Co. v. Leverton, 314 Ill. App. 3d 1080, 1083 (2000)). Under established legal principles, the court may amend the jury verdict

“only when the defect is one of form, rather than substance. [Citation.] A trial court should not amend a verdict in order to reach a determination that the court believes the jury ought to have made, and an amendment must reflect only what the jury clearly intended the verdict to be.” Crowell v. Parrish, 159 Ill. App. 3d 604, 608-09 (1987).

The choice of verdict form B shows that all jurors agreed that Sarrafi’s negligence proximately caused damages to Sofia. In light of the jurors’ use of the proper verdict forms for finding no liability on all other counts, the choice of verdict form B does not appear to result from inadvertent error or confusing trial court instructions. The jurors also agreed, by signing a verdict awarding $0 in damages, that plaintiffs did not prove that Sarrafi negligently caused their damages.

The verdict form here, like the verdict in Tindell v. McCurley, 272 Ill. App. 3d 826 (1995), is simply inconsistent. In Tindell the jury returned a verdict in favor of the plaintiff but assessed $0 in each of the four separate categories of damages, for a total award of $0. The appellate court said:

“In this case, because of the nature of the jury’s verdict there is no way — other than an absolute guess — for this court to say whether the jury found liability. The verdict form was filled in by the jury with zero damages for four separate categories of damages despite uncontradicted evidence plaintiff suffered some damages from the accident. However, to say the jury meant to find for defendant on the issue of liability, we would have to assume the jury ignored the instructions both as to which verdict form to use and that there would be no occasion to consider damages if it found for defendant.” Tindell, 272 Ill. App. 3d at 830-31.

The court ordered a new trial on all issues.

Defendants rely on Kleiss v. Cassida, 297 Ill. App. 3d 165 (1998), as support for the trial court’s interpretation of the verdict here. In Kleiss the jurors asked whether a finding of negligence required assessment of any minimum damages. The court answered “No.” Kleiss, 297 Ill. App. 3d at 175. The jury returned the verdict in favor of the plaintiffs, but assessed no damages. The appellate court found that the jury’s question and the court’s response led to the conclusion that the jurors found in favor of the defendant on the issue of proximate cause, but they used the wrong verdict form.

Here, by contrast, another reasonable hypothesis defendants suggested could explain the verdict. After hours of deliberations the jurors finally reached a compromise: those who believed Sarrafi caused no damages signed a verdict finding Sarrafi liable for damages, in exchange for the other jurors’ signature on a formal assessment of $0 in liability; those who believed Sarrafi negligently harmed Sofia gave up the assessment of damages in exchange for the other jurors’ signature on a formal finding of negligence and liability. Here, as in Manders v. Pulice, 44 Ill. 2d 511, 518-19 (1970), the hypothesis that the jurors intentionally disregarded the court’s instructions and returned an inconsistent verdict as a result of compromise is at least as defensible as the hypothesis that the jury mistakenly signed the wrong verdict form. Because the judgment in favor of Sarrafi did not reflect the clear intention of the jury — this internally inconsistent verdict expressed no intention clearly — the trial court abused its discretion by substituting the verdict the court believed the jury should have reached for the verdict the jury actually rendered.

Following Manders and Tindell, we hold that the irreconcilably inconsistent verdict requires reversal of the judgment in favor of Sarrafi on count I and remand for retrial on liability as well as damages.

II

We expect issues concerning the admissibility of Sarrafi’s testimony about his notes to recur on retrial. Plaintiffs argue that the trial court violated the Act by permitting Sarrafi to testify, from his notes, about his conversations with Sofia. The Act provides:

“In the trial of any action in which any party sues or defends as the representative of a deceased person or person under a legal disability, no adverse party *** shall be allowed to testify on his or her own behalf to any conversation with the deceased or person under legal disability or to any event which took place in the presence of the deceased or person under legal disability, except in the following instances:

(a) If any person testifies on behalf of the representative to any conversation with the deceased or person under legal disability or to any event which took place in the presence of the deceased or person under legal disability, any adverse party or interested person, if otherwise competent, may testify concerning the same conversation or event.

(c) Any testimony competent under Section 8 — 401 of [the Code of Civil Procedure], is not barred by this Section.” 735 ILCS 5/8—201 (West 2000).

The trial court found that subsection (c) justified admission into evidence of Sarrafi’s notes and his testimony about his conversations with Sofia. Section 8—401 of the Code of Civil Procedure (Code) provides:

“Where in any action or proceeding, the claim or defense is founded on a book account or any other record or document, any party or interested person may testify to his or her account book, or any other record or document and the items therein contained; that the same is a book, record, or document of original entries, and that the entries therein were made by himself or herself, and are true and just; *** and thereupon the account book and entries or any other record or document shall be admitted as evidence in the cause.” 735 ILCS 5/8—401 (West 2000).

Thus, the section makes an account book or a similar document admissible in evidence with appropriate testimony only if a claim or defense is “founded” on the document. The parties have not cited any cases interpreting the term “founded” in this section of the Code, and our research has not uncovered any such cases.

When a word occurs in several parts of a statute, it “will be given a generally accepted and consistent meaning where legislative intent is not clearly expressed to the contrary.” Baker v. Salomon, 31 Ill. App. 3d 278, 281 (1975). Sections of a code should be construed together to give a harmonious meaning to the code as a whole. Lemont-Bromberek Combined School District No. 113(a) v. Walter, 279 Ill. App. 3d 847, 850 (1996). The court discerning the legislative intent in the use of a term may look to the use of that term in other statutes concerning similar subjects. In re Application for Tax Deed, 311 Ill. App. 3d 440, 444 (2000); see Pettigrove v. Parro Construction Corp., 44 Ill. App. 2d 421, 425 (1963).

Section 2—606 of the Code also concerns claims or defenses founded on documents. That section provides:

“If a claim or defense is founded upon a written instrument, a copy thereof, or of so much of the same as is relevant, must be attached to the pleading as an exhibit or recited therein, unless the pleader attaches to his or her pleading an affidavit stating facts . showing that the instrument is not accessible to him or her.” 735 ILCS 5/2—606 (West 2000).

Several cases have interpreted the statutory limitation to claims “founded” on the written instrument. In McCormick v. McCormick, 118 Ill. App. 3d 455 (1983), the beneficiary of a trust sued the trustees for breach of trust. The beneficiary attached to his complaint the document creating the trust as well as numerous documents concerning the transactions the trustees directed for the trust. The beneficiary contended that the transactions so documented wasted trust assets. The appellate court held that the trust document qualified as a writing on which a claim was founded, but the other documents did not so qualify. The court decided that the other documents were “example[s] of the evidence supporting plaintiffs allegations” (McCormick, 118 Ill. App. 3d at 460), rather than documents providing the foundation for the plaintiffs claims. McCormick, 118 Ill. App. 3d at 460-61.

The court similarly distinguished evidentiary documents from documents upon which a claim is founded in Garrison v. Choh, 308 Ill. App. 3d 48 (1999). In that case the plaintiff sued a doctor for medical malpractice and attached to the complaint the report of a health professional who found that the plaintiff had reasonable grounds for the malpractice suit. Section 2—622 of the Code required the attachment of such a report to the malpractice complaint. 735 ILCS 5/2—622 (West 1992). The plaintiff argued that the malpractice claim was founded on the report, within the meaning of section 2—606.

The appellate court rejected the plaintiffs argument. The court noted that section 2—606 applied primarily to instruments like contracts, and the court distinguished documents that constituted evidence supporting the allegations of the complaint. Garrison, 308 Ill. App. 3d at 53. The court held that the plaintiffs “claim was not founded on the report; it was founded on the acts of the defendant which were alleged to have been negligent.” Garrison, 308 Ill. App. 3d at 54.

Sarrafi’s medical notes similarly do not form the foundation for his defense of contributory negligence. Sarrafi founded his claim on Sofia’s allegedly unreasonable refusal to follow medical advice; the notes provided evidence of the refusal. The notes here do not operate like a contract, a trust agreement, or an account book as the basis for the claim.

The trial court relied on Herron v. Anderson, 254 Ill. App. 3d 365 (1993), as support for the decision to allow Sarrafi’s notes into evidence. The trial court in Herron permitted the plaintiff to introduce into evidence certain medical records, but refused to redact portions the plaintiff wished to exclude. The appellate court held that the rule of completeness justified the court’s decision not to redact parts of the medical records the plaintiff introduced. Herron, 254 Ill. App. 3d at 375-76. The court then, in dicta, added that all medical records are always admissible, as business records, pursuant to subsection (c) of the Act and section 8 — 401 of the Code. Herron, 254 Ill. App. 3d at 376.

We cannot agree with the dicta from Herron. Section 8 — 401 of the Code, and the Act, do not provide generally for the admissibility of business records or medical records. Section 8—401 expressly limits its reach to documents upon which a claim or defense is “founded.” Herron presents no reason for reading this requirement out of the statute. The dicta in Herron does not mention the phrase limiting the statute to cases in which “the claim or defense is founded on” a document. 735 ILCS 5/8 — 401 (West 2000).

Commentators have long criticized the Act. See, e.g., N. Kahn, Let’s Give the Dead Man’s Statute a Decent Burial, 55 Ill. B.J. 430 (1967); M. Barnard, The Dead Man’s Act Rears Its Ugly Head Again, 72 Ill. B.J. 420 (1984). Nonetheless, any change should come from the legislature, and not under the guise of judicial construction of the statute. The Act, as enacted by the legislature, incorporates section 8—401 of the Code, limiting the exception to documents on which a “claim or defense is founded.” Not all business and medical records qualify as foundations of claims, even when they may constitute evidence supporting such claims. McCormick, 118 Ill. App. 3d at 460-61. Sarrafi’s notes here are evidence supporting his defense, but they do not qualify as documents upon which the defense is founded. Accordingly, we find that exception (c) in the Act does not justify the introduction into evidence of Sarrafi’s notes, nor does it justify the decision to permit Sarrafi to testify from those notes about his conversations with Sofia.

Plaintiffs on appeal directly address only the application of the Act to notes and testimony regarding the office visit on May 28, 1996, and therefore we confine our review to those notes. We observe, however, that the court at trial allowed testimony about other office visits. The court on remand should revisit the rulings on the admissibility of such evidence in light of the analysis presented herein.

Defendants argue that exception (a) to the Act warrants the introduction into evidence of Sarrafi’s notes. Under exception (a), if the representative of a person under a legal disability introduces evidence of a conversation between the person under a disability and another person, any party adverse to the representative may testify about the same conversation. When plaintiffs called Sarrafi as a witness, they narrowly confined their questions to acts he performed on or after June 3, 1996. They did not in any manner refer to his conversations with Sofia on May 28, 1996. The court permitted Sarrafi to testify about his notes of that conversation on defendants’ examination of Sarrafi. Because plaintiffs had not introduced any evidence about the conversation, subsection (a) of the Act cannot justify the admission into evidence of Sarrafi’s testimony about his notes. See In re Estate of Deskins, 128 Ill. App. 3d 942, 952-53 (1984); compare Haist v. Wu, 235 Ill. App. 3d 799, 818 (1992).

Plaintiffs’ later examination of their expert concerning Sarrafi’s notes, brought in response to Sarrafi’s improper testimony, did not serve to waive retrospectively the timely objection to Sarrafi’s prior testimony. Compare Hoem v. Zia, 159 Ill. 2d 193, 198-201 (1994) (the plaintiffs prior elicitation of interpretation of notes opened the door for the defendant’s testimony concerning the subject of the notes).

Neither subsection (a) nor subsection (c) of the Act warranted the introduction into evidence of Sarrafi’s testimony concerning the notes he dated May 28, 1996. On remand, the trial court must exclude the notes and testimony about the conversations between Sarrafi and Sofia, unless plaintiffs first introduce evidence concerning the notes or conversations.

III

Plaintiffs raise no challenge to the judgment entered on the jury’s verdict in favor of Sarrafi and HCSC on counts III and IV of the complaint. But Mooshi asks this court to reverse the judgment entered against him on count II, the claim for loss of consortium based on the facts stated in count I of the complaint.

The trial court should have excluded from evidence Sarrafi’s testimony about his notes of his conversations with Sofia on May 28, 1996. The jury’s inconsistent verdict on count I shows that the jury found the evidence of liability closely balanced. We, too, find that evidence closely balanced. As Mooshi’s proof of liability on count II matched the proof on count I, evidence regarding liability on count II is also closely balanced.

Sarrafi made Sofia’s compliance with his advice a major issue in the case, arguing that any failure to communicate with her had no causal effect because she would not have accepted treatment anyway. The admission into evidence of Sarrafi’s notes and his testimony about his notes therefore constitutes reversible error. See Alvarado v. Goepp, 278 Ill. App. 3d 494, 497 (1996). We reverse the judgment in favor of Sarrafi on count II and remand for a new trial.

IV

Plaintiffs also challenge the trial court’s decision to enter judgment notwithstanding the verdict in favor of HCSC. First, plaintiffs argue that the evidence could support a verdict against HCSC based on the theory of implied authority. Our supreme court, in Petrovich v. Share Health Plan of Illinois, Inc., 188 Ill. 2d 17 (1999), established the applicable parameters:

“One context in which implied authority arises is where the facts and circumstances show that the defendant exerted sufficient control over the alleged agent so as to negate that person’s status as an independent contractor, at least with respect to third parties. *5$: $}:

* * *

No precise formula exists for deciding when a person’s status as an independent contractor is negated. Rather, the determination of whether a person is an agent or an independent contractor rests upon the facts and circumstances of each case. Merlo v. Public Service Co., 381 Ill. 300, 319 (1942). *** Facts bearing on the question of whether a person is an agent or an independent contractor include ‘the question of the hiring, the right to discharge, the manner of direction of the servant, the right to terminate the relationship, and the character of the supervision of the work done.’ Merlo, 381 Ill. at 319. The presence of contractual provisions subjecting the person to control over the manner of doing the work is a traditional indicia that a person’s status as an independent contractor should be negated.” Petrovich, 188 Ill. 2d at 42-47.

In Petrovich the plaintiff presented evidence that the defendant controlled the doctor by providing financial incentives that could reduce the doctor’s income if he ordered too many tests or hospitalizations. Petrovich, 188 Ill. 2d at 47-48. The defendant responded that the doctor had earned a constant monthly salary. Our supreme court answered that the lack of any actual reduction in the doctor’s salary did not insulate the defendant. “Whether control was actually exercised is not dispositive in this context. Rather, the right to control the alleged agent is the proper query, even where that right is not exercised.” (Emphasis in original.) Petrovich, 188 Ill. 2d at 48.

The defendant in Petrovich performed a “ ‘quality assurance review’ ” (Petrovich, 188 Ill. 2d at 49), and it reserved the right to terminate the participation in the network of any physician who had cared for patients in ways of which the defendant disapproved. The defendant also required its doctors to obtain its approval for all referrals to specialists. The court found this evidence sufficient to survive summary judgment on the issue of implied agency. Petrovich, 188 Ill. 2d at 50.

Plaintiffs here similarly presented evidence of financial incentives. HCSC paid Holy Family, rather than any of the individual doctors affiliated with Holy Family, and Holy Family in turn paid the individual doctors. Although HCSC’s financial incentives directly affected only Holy Family, those incentives affected the total amount available to the doctors affiliated with Holy Family, and thereby provided indirect incentives for the doctors to reduce the number of tests performed and hospitalizations ordered. Plaintiffs also presented evidence that HCSC performed a quality assurance review and that it could remove from its network any physician providing care in ways HCSC disapproved.

The most significant difference between this case and Petrovich is that the defendant in that case required its doctors to obtain prior approval for all referrals, while HCSC did not require approval for referrals here. We do not find this difference sufficient to remove from the jury’s determination the question of whether HCSC retained substantial control over Sarrafi. Because the inquiry into control is fact-intensive, and reasonable persons could differ about the implications of those facts regarding control, we reverse the decision to grant HCSC judgment notwithstanding the verdict on the issue of implied authority.

Petrovich also sets forth the standards applicable for proof that a doctor appeared to act as an HMO’s agent.

“To establish apparent authority against an HMO for physician malpractice, the patient must prove (1) that the HMO held itself out as the provider of health care, without informing the patient that the care is given by independent contractors, and (2) that the patient justifiably relied upon the conduct of the HMO by looking to the HMO to provide health care services, rather than to a specific physician. ***

* * *

The element of justifiable reliance is met where the plaintiff relies upon the HMO to provide health care services, and does not rely upon a specific physician. This element is not met if the plaintiff selects his or her own personal physician and merely looks to the HMO as a conduit through which the plaintiff receives medical care.” Petrovich, 188 Ill. 2d at 33-38.

Here, Herminda testified that she recommended the BlueAdvantage HMO, rather than the PPO or traditional insurance, partially in reliance on HCSC’s representations concerning the quality of doctors participating in the network. Herminda recommended keeping Sarrafi as the family’s primary care physician because he participated in the network and because of the family’s established relationship with him. Mooshi testified that he relied on Herminda’s recommendation when he enrolled in the HMO.

Even if we view this evidence as sufficient to show that Mooshi indirectly relied on HCSC’s representations when choosing the HMO rather than the PPO or traditional insurance, the record lacks evidence that the choice of the HMO made any difference for the choice of Sarrafi as the primary physician. Had Mooshi chosen the PPO or traditional insurance, the family still would have had the option of continuing as Sarrafi’s patients. In view of the evidence that Sofia went to Sarrafi for care at least once without insurance coverage, the record presents especially strong grounds for concluding that Sofia would have continued to receive care from Sarrafi under either the PPO or traditional insurance, both of which would have covered Sarrafi’s treatment of her. The record lacks evidence that Sofia would have chosen a different doctor if not for the choice of the HMO rather than the PPO or traditional insurance. That is, Sofia did not rely on representations HCSC had not yet made about its HMO when she first chose to accept treatment from Sarrafi, and Mooshi’s decision to accept the HMO did not affect the continuation of treatment from Sarrafi, who remained an option as the primary care physician under both alternative kinds of insurance. Accordingly, the record cannot support the conclusion that Mooshi and Sofia relied on HCSC’s representations about its HMO when Sofia sought treatment from Sarrafi. Without evidence of reliance, no verdict in plaintiffs’ favor on the issue of apparent agency could ever stand.

CONCLUSION

Plaintiffs presented sufficient evidence to raise a triable issue concerning whether Sarrafi’s negligence proximately caused Sofia’s injury. Experts testified that Sofia’s risk of stroke without appropriate medication was two to three times as great as the risk with such medication. Plaintiffs presented evidence that any reasonable person, including Sofia, would have accepted the medication once confronted with the EKG showing a mass, with a high risk of embolization, lodged in her heart. Thus, the negligent failure to contact Sofia and inform her of the results of the EKG might be found to be a proximate cause of her injury.

The record suggests that the jurors may have reached a compromise verdict when they returned a form finding against both defendants on count I, but assessing $0 in damages. The trial court correctly denied the motion for new trial on damages only in light of the appearance of compromise. But the court erred by amending the verdict to favor the position of the jurors who sought entry of a verdict in favor of defendants. The record presents no basis for concluding that the assessment of $0 in damages was somehow more unanimous than the finding that Sarrafi, as HCSC’s agent, negligently caused damages to Sofia. The inconsistent verdict requires reversal and remand for retrial.

Sarrafi’s notes provided evidence supporting his affirmative defense of contributory negligence, but the defense was not “founded on” the notes within the meaning of section 8—401 of the Code, incorporated by reference into the Dead-Man’s Act. Neither did plaintiffs open the door for admission of the notes. As no exception justifies admission into evidence of Sarrafi’s notes regarding his conversation with Sofia, the Act precludes the use on remand of the notes and testimony about the conversation of May 28, 1996, unless plaintiffs open the door to such testimony by introducing evidence concerning that conversation. The inadmissible evidence may have affected the verdict on count II, so the judgment in favor of defendants on that count is also reversed and remanded.

Plaintiffs did not present evidence that could support a finding that they justifiably relied on the appearance that Sarrafi acted as HCSC’s agent. The evidence of implied agency, however, raises an issue for the trier of fact. The judgments in favor of both defendants on counts I and II are reversed and the cause is remanded for proceedings in accord with this opinion.

Reversed and remanded.

GORDON, EJ., and O’MALLEY, J., concur.