JUSTICE GORDON
delivered the opinion of the court:
Defendant Michael Hill appeals his conviction of first degree murder (720 ILCS 5/9 — 1(a)(1) (West 1994)) by the circuit court of Cook County following a bench trial in August 1995. He argues his conviction must be reversed and remanded for a new trial because he did not receive a fitness hearing despite the fact that he was taking psychotropic medication at the time of the trial. In addition, he argues the court’s rejection of his insanity defense was against the manifest weight of the evidence and his conviction should be reversed and a finding of not guilty by reason of insanity entered.
In April 1993, defendant lived with Marguerite Hill, his wife, and Tracy Hill, his adult daughter. On the evening of April 4, Tracy returned home from work at some time after midnight. When she walked to her parents’ bedroom, she saw her father, fully clothed, sitting on the floor by the bed, cradling Marguerite’s head in his lap. Marguerite was clad only in panties. Tracy immediately backed up and apologized, but received no response. She went to her room and went to sleep.
Defendant woke his daughter at approximately 4:30 a.m. on the morning of April 5, and told her to call 911, but would not tell her why. As the house phone could not make outgoing calls, Tracy got dressed and ran down the street to make the call. When Tracy went into her parents’ bedroom after making the call, she saw her father sitting on the floor and her mother lying next to him with her head in his lap, covered by a blanket or sheet. When Tracy asked defendant what was wrong and tried to look at her mother, defendant leaned down over her so Tracy could not see her face. She identified a necktie found near the victim as one she had given her father. After the paramedics arrived, Tracy examined the windows and doors of the house, looking for signs of forced entry; she did not see any.
Tracy testified that defendant had received psychiatric treatment before April 5, 1993. In 1992 he had received inpatient treatment at Jackson Park Hospital after a suicide attempt. Later in 1992 he was admitted to the Veteran’s Administration Hospital (VA) for approximately a month, after which he received some outpatient treatment through the VA. She had not noticed any changes in defendant’s behavior before either of the two admissions for treatment. She did not remember him being admitted for treatment on any other occasions. She did not state why he had been admitted.
When paramedics responded to the 911 call, they found defendant sitting on the floor of the bedroom with the victim lying beside him, her head cradled in his lap. Defendant did not respond to their questions why they had been called or what was wrong. When the paramedics examined the victim they discovered that she had been deceased long enough for rigor mortis to set in, which took approximately six to eight hours.
The paramedics summoned the police. When the police arrived, around 5:30 a.m., defendant initially failed to respond to their questions. After a while, he stated that he did not remember what had happened. Several detectives arrived at the scene at approximately 6 a.m. Defendant told Detective Edmond Leracsz that he had been watching television the night before, and he had developed a headache, for which he took some aspirin. The aspirin did not help, and the victim offered to rub his temples. At this time the victim was fully clothed. The next thing defendant remembered was being awakened by the alarm clock at 4:30 a.m., when he went to wake his daughter. Defendant repeated this version of events when he spoke with some other detectives at the police station. Detective Leracsz inspected the windows and doors of the house, looking for signs of forced or attempted forced entry, but found no such evidence.
On April 6 defendant spoke with Assistant State’s Attorney Peggy Chiampas at the police station. He repeated the same version of events to her until she asked him what he remembered next after watching television. At that point he stated that he wanted a lawyer, and Chiampas terminated the interview. She described him as being very polite and alert. He looked at her when he spoke, and when she asked questions he responded promptly and appropriately. He did not complain about how he felt.
Deputy Medical Examiner Dr. Mitra Kalelkar conducted the postmortem examination of decedent. She testified that in her opinion the decedent had been the victim of a homicide by strangulation. The parties stipulated, that if called, Cecilia Doyle would have testified that tests performed on bloodstains on the necktie found next to the victim revealed that the blood on the tie was consistent with the victim’s blood.
The State rested, and the defense’s motion for a directed verdict was denied. The defense proceeded to call its first witness, Dr. Henry Conroe. Dr. Conroe, a psychiatrist, first met the defendant in January 1994. He interviewed him two subsequent times as well, for a total of approximately five hours. He formed an opinion that in April 1993 defendant was suffering from a major depressive episode with mood incongruent and psychotic features. He did not only rely on his sessions with defendant in forming his opinion; he also considered an interview that he conducted with defendant’s friend, Phillip Schwartz; police reports of the homicide; records from the VA and Jackson Park; psychological testing on defendant in May 1993 at the Department of Corrections; brief telephone contact with defendant’s treating doctor, Dr. Earl; psychological testing by Dr. Marva Dawkins; and a report prepared by Dr. Henry from the Psychiatric Institute of Cook County.
Conroe testified that in addition to his depression, defendant had “persecutory ideation.” While at the VA hospital he had psychoses centering around his wife, involving a conspiracy in the police department because a police officer supposedly had an affair with his wife. Conroe was aware that there actually had been problems with defendant’s wife’s infidelity and admitted that defendant’s belief that his wife had been having an affair with a Chicago police officer was not psychotic. Conroe testified that defendant told him that issues of his wife’s fidelity no longer concerned him. He no longer believed the entire Chicago police department knew his wife was having an affair, as he once had.
Conroe testified that defendant was admitted to Jackson Park Hospital in April 1992, after taking an overdose of antidepressant medication. He was suicidal and had been for about a year since he had lost his job. He was not eating or sleeping well, felt despondent and could not find a job. He was admitted to the VA in May 1992, shortly after he signed himself out of Jackson Park. Conroe initially testified that defendant’s diagnosis at the VA was that he suffered from major depression with psychotic features. On cross-examination he admitted that the diagnosis was actually a single episode of major depression. Defendant was initially admitted to the VA for inpatient treatment, then switched to outpatient treatment before his discharge. He stopped attending outpatient sessions in November 1992.
Defendant had been taking doxepin (Sinequan), an antidepressant, and Stelazine, an antipsychotic, in late 1992 and early 1993. These medications had been prescribed for him during his treatment at the VA. Doxepin should have helped defendant maintain his energy level and mood, help him sleep and eat and concentrate. Stelazine should have helped him keep any paranoid thinking or persecutory ideation in check. Defendant told Conroe that he had ceased taking the medication two to three weeks before his wife’s murder. Conroe believed that there would be at least some or possibly a full return of symptoms within two to four weeks after ceasing taking the medications.
Conroe believed that at the time of the crime, defendant was suffering from major depression with mood incongruent psychotic features, and he was unable to appreciate the criminality of his behavior and unable to conform his conduct to the requirement of the law. Conroe testified that the defendant was still under treatment at the VA hospital at the time of trial, and from Conroe’s review of the records to date from the VA, it appeared defendant was still receiving Zoloft (misspelled in the record as “solarlov”), an antidepressant, at the time of trial. Conroe admitted that Dr. Henry’s report indicated that the defendant had told him that he was taking his medication at the time of the crime. However, Conroe maintained that his opinion as to whether defendant was sane at the time of the crime would not change even if defendant had been taking his medication at the time. He noted that defendant was deteriorating despite being on doxepin, which indicated the dose was not sufficient or it was not the right medication. He admitted that the only sources of information for defendant’s deterioration were defendant himself and Schwartz. However, Conroe maintained that whether defendant was on medication at the time of the crime meant nothing to him in terms of his opinion of defendant’s sanity at the time.
Conroe admitted that defendant had been taken off Stelazine on September 14, 1992, while still an inpatient at the VA. The records reflected that defendant had denied any paranoid or bizarre behaviors thereafter. The Stelazine was never restored according to the records. There was no evidence in the records from the VA of any bizarre behavior or physical acting out after that date; there were no physical symptoms at least until October 22, 1992. However, Conroe noted that defendant was still taking antidepressants during this time, and adequate treatment of his depression would lessen his psychotic features. Conversely, removal of the antidepressant would accelerate an onslaught of psychotic features. Conroe noted that it was shortly after the Stelazine was discontinued that defendant began missing appointments and not attending the VA clinic. This was also around the same time Schwartz described his behavior change.
Conroe testified that Phillip Schwartz told him that he and defendant had met when both were inpatients at the VA. Their relationship had continued after defendant’s release from inpatient treatment. Defendant had helped Schwartz get a job at Goodwill around Christmas time of 1992, and Schwartz had become concerned about defendant becoming more depressed and not being his usual self in the way he related to people. He spoke with the decedent in late December 1992, and she agreed defendant was becoming depressed and needed to return to the VA. Conroe was not aware of defendant returning to the VA at that time, although he knew that defendant did continue his medication until mid-March 1993.
Conroe believed that Schwartz was not an inpatient at the time they spoke. He did not know what Schwartz’s prior diagnosis was and did not know if he was taking psychiatric medication when he spoke with him. He stated that Schwartz did not seem bizarre or profoundly depressed, and he was able to give a coherent account and answer questions relevantly.
Conroe estimated he had examined about 150 defendants and testified in about 50 trials. All but one of the examinations were at the request of the defense. He admitted co-authoring an article entitled “Sanity Defense — A Practical Guide” with a member of the American Civil Liberties Union. Although he maintained that major depression was a “major mental disorder,” when he was asked if it was a psychosis, he admitted that there was no evidence of psychoses when defendant admitted himself to Jackson Park. When defendant signed himself out of Jackson Park, there were no criteria to certify for involuntary treatment. Conroe admitted that insanity is difficult to establish if the illness itself is not a psychosis.
Conroe stated that defendant had given him a full life history, with numerous details, organized in a coherent fashion. He mentioned events related to his growing up, his education, employment, and alcohol use or nonuse, among other things. However, defendant had a gap in his. memory surrounding the murder. Conroe admitted defendant had no history of memory gaps or amnesia, and he was aware of no organic basis for defendant’s loss of memory. He also admitted that depression would not necessarily impact one’s memory, and it was possible that defendant was lying about his loss of memory. Conroe countered, however, that people sometimes block out memories too emotionally intense for them to handle, and he had noticed that defendant tended to shy away from dealing with situations involving strong emotions.
Conroe also admitted that insanity is often manifested through bizarre behavior. He admitted that he had examined the police reports and found no indication therein of any bizarre or unusual behavior on defendant’s part.
Conroe stated that he spoke with defendant’s son briefly but did not speak with defendant’s daughter, Tracy. He admitted that she would have been able to tell him about the defendant’s mental condition, but he would not admit she would have had better knowledge of defendant’s mental condition than anyone else.
Conroe admitted that sexual dysfunction (one of the issues for which defendant was being treated at the VA) could itself be a source of depression for adult males, as could job loss. This did not mean that as a result of either of these two occurrences one could not conform one’s behavior to the law or appreciate the criminality of conduct. However, Conroe differentiated between “being depressed” and suffering from the medical and psychiatric illness of major depression, with which he had diagnosed defendant.
Conroe’s opinion was in part based on tests administered to defendant by Dr. Dawkins. One of those tests was the Minnesota Multiphasic Personality Inventory (MMPI). Conroe admitted that there were certain scales “built into” the MMPI “as a sort of truth detector.” In this case some of the results raised questions about the validity of the test. However, Conroe did not believe the test was invalid. He explained that it was common to see elevated scores when someone in emotional distress took the test.
Dr. Marva Dawkins, who administered the tests, was the defense’s next witness. Dawkins, a clinical psychologist, was also called as an expert witness on defendant’s sanity. Dawkins’s training is in forensic psychology. She met with defendant in July 1994, at the request of the defense, for approximately three hours. Before she met with him he had been given a battery of tests through her office: the mental status checklist for adults; a symptoms checklist (the SCL 90); the Modified Michigan Alcoholism Screening Test (the Modified Mass); the Bender visual motor test; the Gestalt Test; the Shipley Institute of Living Test; the personal history checklist for adults; and the Minnesota Multiphasic Personality Inventory 2 (MMPI 2). She also reviewed records from defendant’s admissions to Jackson Park and the VA.
Dawkins concluded that defendant suffered from delusional disorder, major depression with psychoses, and paranoid personality disorder. She believed he had been suffering from delusional disorder at the time of the crime, and he was therefore able to neither appreciate the criminality of his behavior nor conform his conduct and behavior to the requirements of the law. Between the time Dawkins prepared her report and trial she reviewed a report prepared by Dr. Henry of the Psychiatric Institute and medical and psychological records from the VA hospital; her opinion had not changed as a result of the additional materials.
Dawkins admitted that defendant had talked “almost non-stop” during her interview with him, and he was able to organize his thoughts and spoke clearly and coherently to her. He gave her a number of details, including names and dates, although Dawkins noted that his date memory was clearer in the last three to four years. She admitted that almost all of the significant dates he told her about involved April 5.1 For instance, he told her that twice in his life he had been fired on April 5. However, although Dawkins found it unusual that all of the traumatic events in defendant’s life occurred on his birthday, she did not attempt to verify this fact. The important thing to her was defendant’s psychological state — not “reality,” but defendant’s reality.
Dawkins believed that defendant’s memory loss during the murder was triggered by a traumatic experience. Specifically, she believed it was triggered by defendant being told that he was going to be fired from Goodwill on his birthday. Although not certain, she thought it was possible that defendant had also experienced a dissociative reaction when he lost his job on April 5 the first time. She also thought it was possible he had experienced one when he lost his job on April 5 the second time, although there was no record thereof. He did describe himself as having experienced an “emotional [sic] induced physiological breakdown.”
Dawkins stated that her opinion would not change if it were not true that defendant was going to lose his job. The important thing was defendant’s perception that he was going to lose his job. She noted that defendant “lived in a world that was filled with delusions that were not true in reality and he acted upon these delusions.” Dawkins admitted that there was no indication of outside hallucinations in any reports and that defendant had not reported any command hallucinations. Dawkins stated that she could not answer the question whether it was equally possible that defendant’s actions were volitional, as opposed to the result of a dissociative reaction. She explained this question went to “the heart of a philosophical question of what is volitional and what is functionally mentally ill.”
Dawkins admitted that certain sections in the MMPI — the L, F and K scales — could be described as validity scales and are included in order to alert the examiner to the possibility that the subject is faking his responses. Dawkins admitted that she had indicated in her report that defendant’s responses raised questions about the overall scale validity. She admitted that the whole test result could be questioned based on the indicators. However, she noted that there could be a variety of reasons for the responses. One reason is that the subject is lying, but another possibility is the subject’s emotional distress. Dawkins noted that she had not interpreted the MMPI “in the blind”; she used other data, including the other tests, her face-to-face interview, and her knowledge of defendant’s history to help her interpret the test. She concluded that the elevated scales were not indicative of false answers but rather of emotional and mental disturbance.
Dawkins admitted that if an intelligent individual took the tests she administered to defendant with the intent to feign a psychiatric illness, he could do so. However, she was aware of the problem of malingering and did not believe that defendant fit the profile of a malingerer.
In rebuttal the State called Dr. Stafford Heniy. Dr. Henry, a forensic psychiatrist, was an employee of the circuit court of Cook County and had been so employed since July 1993. Henry examined defendant in September and October 1994, for a total of approximately five hours. One of the first things he noticed about defendant was that he had a guarded and suspicious demeanor and was wearing sunglasses indoors. Henry found the fact that defendant wore sunglasses during the evaluation significant, because it did not allow Henry to see his facial expressions; this led Henry to believe that perhaps defendant had “something that he did not want me to see.” Henry described defendant as having an excellent command of the English language, but being very deliberate and measured in what he said, his responses having a “somewhat rehearsed quality.” Later in his testimony Henry returned to this, stating that he believed the most remarkable things about defendant were that he was guarded and suspicious, not very forthright, but had an excellent command of English and very deliberate measured speech.
Defendant was able to give Henry a very detailed account of his behavior during the day of the crime, which suggested to Henry that defendant’s intellectual faculties and thought processes were intact. Defendant related a number of details to him regarding the events of the evening of the murder before his wife died, including observations and jokes he and the victim had made about the movie they were watching. However, defendant purported to have no recollection of how his wife died; he had no memory of the events preceding the moment he told his daughter to call 911. Henry testified that defendant told him that he had been taking his medication on the date of the murder; Henry confirmed this information through (unspecified) collateral sources.
Henry conducted a “character logic assessment” of defendant, which is an assessment of how someone sees himself and interacts with the world. Henry saw defendant as being narcissistic and having some degree of antisocial personality disorder. Henry stated that the hallmark of antisocial personality disorder is being manipulative and lacking regard for the truth and for others; he defined narcissism as a sense of entitlement or being above laws or rules.
Defendant’s employment status was among the things that he discussed with Henry. Henry found no evidence suggesting that defendant was to be fired in April 1993. In fact, defendant told Henry that he had just finished a project ahead of schedule and under budget. Henry saw a May 1995 letter from Goodwill which stated that defendant no longer worked there because he had “abandoned his responsibilities” — Henry believed this referred to the time when defendant was in custody after the crime.
Henry saw no evidence that defendant was psychotic, meaning that he was having auditory or visual hallucinations, and stated that “cognitively he was sharp as a tack.” There was no evidence of an organic basis for any memory loss. Henry had a difficult time assessing defendant’s mood because he did not present his history accurately; Henry felt that often defendant said things because they were in his best interest. He admitted that defendant had mentioned feeling like there were people “out to get him” and had recounted one specific episode in 1994 when he thought someone was spiking his drinks in a bar; there was also evidence in prior records that defendant suffered from paranoia.
Before he reached his final conclusion, Henry consulted collateral sources (in addition to the records he had already examined). He spoke with defendant’s two daughters, his son-in-law, and an assistant State’s Attorney. His interviews with the family members lasted approximately an hour and a half.
Henry diagnosed the defendant first as having a history of alcohol dependence. His second diagnosis was “possible” major depressive disorder. His final psychiatric diagnosis was that defendant had unspecified personality disorder with antisocial and narcissistic features. He qualified his diagnosis by stating that it was in part based on information provided by defendant, on whose credibility he did not rely. He stated that this suspicion was corroborated by the results of the MMPI administered by Dr. Dawkins, noting that all of the scales included in the MMPI to detect malingering (the L, F and K scales) were elevated in the results of defendant’s testing.
Henry’s opinion was that on the date of the murder defendant was not suffering from a mental illness and was therefore sane at the time of the crime. He noted that defendant was taking his medication, described himself as being very productive at work, and had been engaging in other activities uncharacteristic of someone severely depressed — joking with his wife and reading one or more books during the day.
Henry criticized on two grounds Dawkins’ determination that defendant was suffering from a dissociative state at the time of the murder. First, he stated that any of Dawkins’ results would be difficult to substantiate because the MMPI results indicated that defendant lied and exaggerated his symptoms. Second, psychiatric illness generally occurs “longitudinally,” meaning that symptoms recur. Here there was no history of defendant entering a dissociative state before the crime or since.
Henry also disagreed with Dr. Conroe’s diagnosis that defendant was suffering from a major depressive episode at the time of the crime. First, as with Dawkins, he felt the opinion had to be discounted because it was so heavily based on information provided by defendant. Second, Henry felt there was no evidence defendant was depressed around the time of the offense. He noted again that defendant had reported having completed a project ahead of schedule and below budget; he had been having a jovial conversation with his wife on the evening of April 4; and he read at least one entire book on that date. This behavior would be inconsistent with someone being in the throes of a major depressive episode.
Henry did admit on cross-examination that elevation on the L, F and K scales of the MMPI could be explained by a severe psychiatric disturbance as well as by malingering. However, he felt that in this case a severe disturbance was not the explanation for the elevation. He was aware that the defense experts had reached a different conclusion and was also aware that the 1992 VA admission report diagnosed him with major depression with psychotic features. However, he felt that since all of these diagnoses were based on information provided by defendant, they were of little value. He noted that he spoke with three other psychologists in his office, and all three agreed that the profile was “highly suggestive” of malingering.
He admitted that defendant had not been taking Stelazine or any other antipsychotic medication at the time of the crime, but his understanding was that the Stelazine had been withdrawn because defendant was no longer psychotic. He discounted the fact that the dose of Stelazine had been increased while defendant was an inpatient at the VA because he attributed it to defendant’s subjective reporting of his symptoms.
When Henry asked defendant why he was wearing sunglasses, he was told that they were for light sensitivity. Henry did not ask defendant who his optometrist was, nor did he make any attempt to verify defendant’s condition of light sensitivity. He also did not verify defendant’s statement that he had just brought in a job at Goodwill ahead of schedule and under budget. Nor did he consult the most recent VA records, since, he explained, his primary concern was defendant’s mental state at the time of the crime. He did, however, consult records through 1994.
At the time Henry spoke with defendant, he was taking Zoloft, an antidepressant. Defendant told him he was taking two other medications as well; one was Calan, a blood pressure medication, and defendant did not recall what the other one was. Before the trial court made its ruling, it sustained an objection made by the defense at trial and stated that, based on Wainwright v. Greenfield, 474 U.S. 284, 88 L. Ed. 2d 623, 106 S. Ct. 634 (1986), it would not consider defendant’s invocation of his Miranda right to an attorney in evaluating his sanity. Nevertheless, the court found defendant guilty of first degree murder and found the defense had not proven defendant was insane at the time of the crime. The court discounted the conclusions of Drs. Conroe and Dawkins as being more or less completely based on what they were told by the defendant. It found Dr. Henry’s testimony much more credible and found defendant was sane.
Defendant appeals, raising two issues. First, he contends his conviction must be reversed and the cause remanded for a new trial because he was taking psychotropic medication at the time of trial but did not receive a fitness hearing. Second, he contends the trial court’s conclusion that he did not prove himself insane was against the manifest weight of the evidence.
ANALYSIS
I. PSYCHOTROPIC MEDICATION AND FITNESS HEARING
Defendant’s first argument is that his conviction must be reversed and the cause remanded because he did not receive a fitness hearing despite the fact he that was taking Zoloft at the time of trial, as Dr. Conroe testified. He argues that the version of section 104 — 21(a) of the Illinois Code of Criminal Procedure of 1963 (Code) which was in effect during his trial (725 ILCS 5/104 — 21(a) (West 1994)) guaranteed-him a fitness hearing because he was receiving psychotropic medication, and he also argues that the only remedy for his not having received such a hearing is to reverse his conviction and remand the cause for a new trial after a proper fitness hearing.
The State does not dispute that defendant was taking psychotropic medication at the time of trial and did not receive a fitness hearing. However, it argues, first, that this court should apply the statute as amended subsequent to defendant’s trial, under which it argues defendant is entitled to no relief. Second, it argues that even if this court applies the original version of the statute, or if we find defendant entitled to relief under the amended statute, the relief to which defendant is entitled is not an automatic reversal of his conviction but, rather, a retroactive fitness hearing.
A. Was defendant entitled to a fitness hearing — What statute applies?
Between the time of defendant’s trial and now, section 104 — 21(a) of the Code has twice been amended by the legislature. Until December 1995, section 104 — 21(a) provided that “[a] defendant who is receiving psychotropic drugs or other medications under medical direction is entitled to a hearing on the issue of his fitness while under medication.” 725 ILCS 5/104 — 21(a) (West 1994). The legislature amended this section effective December 31, 1995, to provide “[a] defendant who is receiving psychotropic drugs under medical direction is entitled to a hearing on the issue of his or her fitness while under medication; however, no hearing is required unless the court finds there is a bona fide doubt of the defendant’s fitness.” 725 ILCS 5/104 — 21(a) (West Supp. 1995).2 The statute was amended effective December 31, 1996, to provide “[a] defendant who is receiving psychotropic drugs shall not be presumed to be unfit to stand trial solely by virtue of the receipt of those drugs or medications.” 725 ILCS 5/104 — 21(a) (West 1996).
Our supreme court has recently made clear that amended section 104 — 21(a) may not be given retroactive effect. See People v. Kinkead, 182 Ill. 2d 316, 333, 335 (1998). The version of the statute that applies is that which was in effect on the date defendant’s right to a fitness hearing vested. People v. Carlson, 293 Ill. App. 3d 984, 986 (1997) (since section 104 — 21(a) protects a constitutional right, it will not apply retroactively if that right has already vested); People v. Cortes, 181 Ill. 2d 249 (1998).3 But cf. People v. Jackson, 294 Ill. App. 3d 1081 (1998) (interpreting Cortes to stand for the unilateral proposition that “the preamended version of section 104 — 21(a) must be applied on direct appeal”). At the latest, a defendant’s right to a fitness hearing vests at the time of sentencing. Carlson, 293 Ill. App. 3d at 988, citing People v. McKay, 282 Ill. App. 3d 108, 114-15, 668 N.E.2d 580, 585-86 (1996). In this case defendant’s trial (which concluded in August 1995) and sentencing (in October 1995) both concluded before section 104 — 21(a) had been amended even for the first time. Accordingly, the pre-December 1995 version of the statute applies, and defendant was entitled to a hearing on his fitness because he was receiving psychotropic drugs.4 725 ILCS 5/104 — 21(a) (West 1994).
B. What is the remedy for defendant not having received a fitness hearing?
The next question is, What is defendant’s remedy for not having received the fitness hearing at the time of trial? Defendant argues his conviction and sentence must be reversed and the cause remanded for a new trial. The State argues that we need merely remand for a retrospective fitness hearing.
Before 1997, the remedy ordinarily would have been to vacate defendant’s conviction and sentence and grant him a new trial, if factual grounds for a hearing existed. See, e.g., People v. Birdsall, 172 Ill. 2d 464, 476, 670 N.E.2d 700, 706 (1996). However, our supreme court has recently clarified that even under the pre-December 1995 version of section 104 — 21(a) it is not always necessary to vacate the conviction of a defendant who did not receive a fitness hearing. Rather, a trial court may hold a retrospective fitness hearing, and if the evidence adduced at such a hearing “ ‘establishes that the defendant did not, in fact, suffer any impairment as a result of his ingestion of psychotropic medication,’ ” defendant is not entitled to a new trial and his conviction may be affirmed. Cortes, 181 Ill. 2d at 276, quoting People v. Neal, 179 Ill. 2d 541, 552 (1997), citing People v. Burgess, 176 Ill. 2d 289, 302-04, 680 N.E.2d 357, 363-64 (1997). The rule announced in these cases applies to cases pending at the time they were announced, including cases pending on direct review, such as the instant case. Cortes, 181 Ill. 2d at 276, citing Neal, 179 Ill. 2d at 552-54.
The decisions in Cortes, Neal, and Burgess establish that a defendant’s fitness may be determined by a retrospective fitness hearing. However, it is not entirely clear in what circumstances such a hearing will suffice. In Neal our supreme court cautioned that “retrospective fitness determinations will normally be inadequate to protect a defendant’s due process rights when more than a year has passed since the original trial and sentencing,” this rule yielding only in “exceptional cases.” Neal, 179 Ill. 2d at 554. More than two years have passed since defendant’s sentencing in this case. Accordingly, the question is what must be found on remand for this to constitute an “exceptional case,” such that defendant need not receive a fitness hearing and a retrial.
Burgess characterized the retrospective hearing as a “case-specific inquiry into the psychotropic drugs administered to this particular defendant.” 176 Ill. 2d at 303, 680 N.E.2d at 363. It stated that courts “should not automatically.assume that every psychotropic drug will inevitably render the person taking it unfit for purposes of trial or sentencing, and we therefore conclude that retrospective hearings are sometimes proper.” Burgess, 176 Ill. 2d at 304, 680 N.E.2d at 364. The court held:
“The evidence in this case, including the prescribing doctor’s testimony [at the retrospective hearing, to the effect that at the dosages he prescribed, none of the medication could have had any effect on defendant’s mental condition even if taken in combination], the judge’s observations [of defendant at trial], and the defendant’s own testimony at trial, compels the conclusion that the defendant was suffering no impairment as a result of his ingestion of psychotropic drugs during the time of his trial and sentencing hearing.” Burgess, 176 Ill. 2d at 304, 680 N.E.2d at 364.
Neal clarified the application of Burgess. It noted that, although in that case approximately 15 years had passed since defendant’s trial and sentencing,
“[t]he passage of time *** is not dispositive [of the question whether a retrospective hearing may be held]. The federal decisions do not establish a bright line rule. Rather than imposing a flat ban on retrospective fitness determinations at delayed post-conviction hearings, they represent an admonition as to the inherent difficulty of retrospectively determining an accused’s competency to stand trial [citation], a principle we well appreciate. [Citation],
Consistent with the United States Supreme Court’s admonition, we cannot dispute that retrospective fitness hearings will normally be inadequate to protect a defendant’s due process rights when more than a year has passed since the original trial and sentencing. In exceptional cases, however, circumstances may be such that the issue of defendant’s fitness or lack of fitness at the time of trial may be fairly and accurately determined long after the fact. In such cases, Burgess will apply, and a defendant will not automatically be entitled to have his original conviction and sentence automatically set aside for a new trial.” Neal, 179 Ill. 2d at 553-54.
Neal went on to state that the situation before it was “directly analogous to that present in Burgess, where, as here, the evidence showed that the medication ingested by the defendant could not have had any effect on his fitness.” Neal, 179 Ill. 2d at 554. The sole evidence to which the court referred in Neal was a psychiatrist’s testimony at a retrospective hearing that at the prescribed dosage the drug defendant was taking was “unlikely” to impair mental abilities and rarely evoked psychotic symptoms and that persons taking the drug “generally improve their cognitive ability.” Neal, 179 Ill. 2d at 547. The psychiatrist in Neal not only testified as an expert in pharmacology, psychiatry and medicine, but was on the board of advisors of the primary manufacturer of the drug which defendant was taking; the court characterized him as having “extensive knowledge of [the drug’s] development, use and chemical properties.” Neal, 179 Ill. 2d at 546.
It would seem after Neal that a retrospective determination of fitness may be upheld if the State introduces expert testimony to the effect that the drug or drugs defendant was taking could have had no adverse impact on his fitness at the prescribed dosage. However, this does not rule out the possibility that a trial court could determine defendant was fit based on other evidence. We note that — although in each case expert drug/dosage testimony was also introduced — our supreme court has cited other evidence in support of retrospective fitness determinations, such as defendant’s testimony and the trial court’s observations of defendant (Burgess, 176 Ill. 2d at 304, 680 N.E.2d at 364), and pretrial fitness examinations which concluded defendant was fit to stand trial, or “fit to stand trial with medication” (Cortes, 181 Ill. 2d at 277). We cannot rule out the possibility that a fitness determination could be made at a retrospective hearing based on evidence other than expert drug/dosage testimony ruling out the possibility of psychotropic effects.5 However, exclusive reliance on other factors will be closely scrutinized, in order that the exception allowing retrospective hearings does not swallow the rule that reversal and remand is the appropriate remedy. We stress that the evidence, of whatever character, must be sufficiently clear that the court may “fairly and accurately” (Neal, 179 Ill. 2d at 553-54) determine defendant’s sanity at the time of trial despite whatever length of time has elapsed since his trial and sentencing.6
Accordingly, we remand for the circuit court to conduct a retrospective hearing on the issue of whether defendant was fit to stand trial. If the court determines at that hearing that this is one of those exceptional cases in which defendant’s sanity at the time of trial may fairly and accurately be determined despite the time intervening, his conviction may be affirmed. If the evidence is inconclusive or suggests that defendant was impaired, his conviction and sentence must be vacated and he must be granted a new trial.
II. SANITY
Defendant also argues that the circuit court erred in rejecting his defense of insanity. He requests that this court vacate his conviction and enter a verdict of not guilty by reason of insanity.
In Illinois all defendants are presumed to be sane. People v. Williams, 265 Ill. App. 3d 283, 289, 638 N.E.2d 345, 349 (1994). The burden is on the defendant to prove insanity by a preponderance of the evidence. 720 ILCS 5/6 — 2(e) (West 1994).7 We will not reverse a trial court’s conclusion regarding sanity unless it is against the manifest weight of the evidence. People v. Johnson, 146 Ill. 2d 109, 128-29, 585 N.E.2d 78, 86 (1991); People v. Baker, 253 Ill. App. 3d 15, 27, 625 N.E.2d 719, 726 (1993). The trial court is free to accept one expert’s testimony over another; experts’ comparative credibility and the weight to be accorded their respective testimony are matters for the trier of fact to determine. Williams, 265 Ill. App. 3d at 289, 638 N.E.2d at 349. The relative weight to be given an expert witness’ opinion is measured by the reasons given for the conclusion and the factual details supporting it. People v. Sojak, 273 Ill. App. 3d 579, 587, 652 N.E.2d 1061, 1067 (1995).
We do not find grounds to reverse the circuit court’s judgment. The expert testimony, set out in detail above, conflicted. The two defense experts differed not only from the State, but from each other, in their diagnoses of defendant. As Dr. Henry observed, the defense experts’ opinions were suspect because of the degree to which they relied on the accuracy of the information provided them by defendant despite indications in the MMPI that defendant was not telling the truth. Also, both defense experts were somewhat discredited on cross-examination. For example, Dr. Conroe stated that his opinion that defendant was insane because of major depression would not change even if defendant was taking his prescribed antidepressant medication at the time of the crime. Dr. Dawkins stated that it did not matter to her whether what she believed was the triggering event for defendant’s dissociative state at the time of the crime had actually happened, because it was defendant’s perceptions, not objective reality, which mattered. The trial court was entitled to consider these peculiarities in determining what weight to accord the experts’ testimony.
Dr. Henry was also impeached to some degree by his failure to investigate certain facts, including defendant’s assertion that he wore sunglasses because of light sensitivity. However, it is in a situation just such as this, when it is not entirely clear who is correct, that deference to the judgment of the trial court is most appropriate. “That the witnesses reached different medical conclusions based upon the same foundational evidence does not mean defendant sustained his burden of proof or that the State failed to carry its burden. *** The contradictory expert opinion presented the trial judge with a classic question of fact.” Sojak, 273 Ill. App. 3d at 588, 652 N.E.2d at 1068. We do not find the conclusion that defendant was sane to have been against the manifest weight of the evidence.
As a final note, our affirming the trial court’s conclusion that defendant was sane does not mean that the trial court is bound to find defendant sane if a retrial is necessary. If the retrospective fitness hearing does not establish defendant’s fitness to stand trial, the court will have to draw its own conclusions regarding his guilt and his saiiity — if he chooses to present that defense — on retrial.
III. CONCLUSION
For the reasons above stated, we reverse and remand to the circuit court for further proceedings consistent with this opinion.
Reversed and remanded.
LEAVITT, EJ., and CAHILL, J., concur.