Summary
holding harmful error exists when there is "reason to believe that an ALJ ignored important evidence"
Summary of this case from Krystal C. v. SaulOpinion
No. 11-1740.
Argued October 4, 2011.
Decided October 21, 2011.
Before RICHARD D. CUDAHY, Circuit Judge JOEL M. FLAUM, Circuit Judge JOHN DANIEL TINDER, Circuit Judge.
Appeal from the United States District Court for the Southern District of Indiana, Indianapolis Division No. 1:09-cv-1429-TWP-MJD.
ORDER
David Walters claims to be disabled by a mental disorder that hinders his concentration and social skills. In denying benefits, an administrative law judge cited several medical reports and records, but did not mention the report of a psychologist enlisted by the Social Security Administration ("SSA") to examine Walters. Unlike other medical providers, the author of that report opined that Walters has poor concentration and poor social skills, which could affect his ability to work. The ALJ needed to explain why he gave no weight to this report, and the Commissioner acknowledges that failing to do so was error. Though the Commissioner insists that the error was harmless, that contention is not persuasive. We thus remand the case to the agency for further review.
Walters applied for disability benefits in April 2005 at the age of 46. He claimed to have been disabled from an unspecified impairment since April 7, 2003, the day he learned that he needed a knee replacement. In his application, Walters reported that he had worked for a construction company in 2002 and 2003 but was unemployed from 1980 to 1986 and "did not have any proof" of earnings he "might have had" for 1990 through 2000. He would later assert on a Social Security Administration questionnaire that he worked only two jobs during the 15 years before he applied for benefits: as a carpenter full time from 1988 to 2000, and at a car wash for two days shortly before he applied for benefits. Records show, however, that Walters had earned a total of $6,911 in 2004 and 2005, and $13,631 in 2006 from working various manual labor jobs.
Walters originally asserted that he suffers from knee pain, bipolar disorder, and schizophrenia, but this appeal concerns only the mental impairments. Medical records show a history of paranoia dating back to 2003. In January of that year the police took Walters to Columbus Regional Hospital because he was a danger to himself, as he feared that people were after him and his house was haunted. He tested positive for methamphetamine and benzodiazepines; he said he was free of street drug use and someone at home had poisoned his food. The hospital staff noted that Walters was on probation because of DUI arrests and appeared to have a chronic problem with substance abuse and dependency. Walters responded well to low-dose antipsychotic medication and was referred to Quinco Behavioral Health Systems upon discharge. At Quinco he met with Dr. Susan Schneider, who prescribed Risperdal and Seroquel for his paranoia and other psychotic symptoms. About a year later, according to a treatment note from Volunteers in Medicine, a general clinic that Walters frequented for his knee problem, he was seen at Quinco for paranoid schizophrenia. He dropped out of treatment there in mid-2004. He also stopped taking Risperdal that year because the drug was causing weight gain.
Walters resumed psychiatric treatment shortly after submitting his April 2005 application for benefits. In May he went to Quinco with several psychological stressors. He was diagnosed with polysubstance and alcohol dependence, but a therapist noted that Walters did not acknowledge any specific mental-health issues. In June, Counselor Joyce Briggs at Volunteers in Medicine reported that Walters was doing better, but easily angered if he believed people were invading his space. Walters told Briggs that he had bouts of depression and had made several suicide attempts, one of them in a jail. He also described a history of hallucinations, though Briggs was skeptical of the claim. Her assessment was paranoid schizophrenia. Later that same month, Walters went back to Quinco to restart his medication. Nurse Susan Kell's diagnostic impression at that time was "Psychotic Disorder NOS, Rule Out Schizophrenia." In July 2005, Walters told staff at Quinco that he was about to lose his job because he could not get along with people at work. Later that month, Dr. Sherman Franz at Volunteers in Medicine assessed Walters's condition as paranoid schizophrenia but acknowledged that this diagnosis might change once the staff became better acquainted with him. Walters told Dr. Franz that an increase in his Risperdal prescription seemed to be helping his symptoms.
Psychologist Karl Evans — at the request of the SSA — more thoroughly evaluated Walters in July 2005. Dr. Evans performed a mental-status examination and diagnosed Walters with Shizoaffective Disorder. Dr. Evans wrote that Walters's immediate rote memory was low, "likely related to poor concentration." The doctor also noted that Walters's attention span was below average but adequate to attend a simple work routine. Dr. Evans reported that Walters's symptoms improve with medication, but even then he experiences significant difficulties with paranoid thinking and poor social skills. Dr. Evans assigned Walters a Global Assessment of Functioning ("GAF") score of 45. That score suggests a "serious impairment in social, occupational, or school functioning (e.g. no friends, unable to keep a job)." AM. PSYCHIATRIC ASS'N, DIAGNOSTIC AND STATISTICAL MANUAL OF MENTAL DISORDERS 34 (4th ed. 2000). Dr. Evans did not draw any conclusion about whether Walters was disabled.
The SSA sent Walters to two more psychologists for examination. In August 2005, Dr. B.R. Horton concluded that Walters was only moderately limited in maintaining social functioning and concentration, persistence, or pace. Dr. Horton's handwriting is barely legible, but there's no dispute that he opined that Walters is capable of performing simple, repetitive tasks if he takes his medications. Then in November 2005, Dr. Richard Karkut diagnosed Walters with depression having mild psychotic features. Dr. Karkut assigned Walters a GAF score of 65, and noted that 70 had been his highest score in the previous year. Those scores are consistent with only mild limitations in occupational functioning. AM. PSYCHIATRIC ASS'N, supra, at 34.
Walters continued treatment at Quinco at least until October 2007. The medical records from late 2005 through 2007 chronicle a varying degree of functioning. In December 2005, Dr. Schneider noted an impression of bipolar disorder but observed that Walters seemed to be "doing quite well" after a knee replacement. In February 2006, however, Walters reported increased depression with some suicidal ideation. Then in March he denied suicidal ideation. Walters described trouble keeping various jobs; he often struggled to keep pace. During his last documented appointment on October 24, 2007, Nurse Kell remarked that Walters "seemed to be worsening as far as depression and other symptoms are concerned." Nevertheless, Quinco physicians consistently rated his GAF at 61. This score suggests mild (but just on the border of moderate) limitations in occupational functioning. AM. PSYCHIATRIC ASS'N, supra, at 34.
Walters explained his job woes during a hearing in front of the ALJ in January 2008. He said that he had earned about $2,200 through a temp agency in 2004, and in 2005 had been earning about the same amount at a fast-food restaurant until he was fired for allegedly hugging a manager in an inappropriate manner. The next year and into 2007, Walters twice was hired for an assembly-line job but was not retained after a 90-day probationary period. He testified that his medications killed his motivation and concentration, but he had told the SSA previously that he was fired from one of the positions because he failed a drug test. Walters quit or was fired from three other jobs in 2007. He had clashes with coworkers at each one: at a dishwashing job because he thought a coworker was disrespecting him, at a catering job because he did not get along with his boss, and at a fast-food job because he thought his Spanish-speaking coworkers were taunting him.
At the hearing, Walters's lawyer highlighted the differences between the reports of Dr. Evans and Dr. Karkut. The ALJ did not comment. Later the ALJ asked a vocational expert ("VE") whether jobs exist for someone with Walters's age, education, and work experience, based on two different hypothetical assumptions. First, he asked the VE to assume that the person could not endure contact with the general public or more than occasional, superficial contact with coworkers, and would be limited to jobs involving routine, repetitive tasks with simple instructions. The VE answered that Walters's past work would be ruled out but that jobs as an inspector or hand packager would be suitable. Second, the ALJ asked the VE to assume that the same person could have no contact with coworkers. She said that no job exists for such a person.
In denying benefits, the ALJ applied the 5-step process required by 20 C.F.R. § 416.920(a) and concluded that (1) Walters had not engaged in substantial gainful activity since the application date, (2) his bipolar disorder and history of polysubstance abuse qualify as severe impairments, (3) these impairments do not collectively meet or equal a listed impairment, (4) he has the residual functional capacity ("RFC") to perform light work, and (5) with this RFC he cannot perform his past work but can meet the requirements of other jobs and thus is not disabled. The RFC determination limits Walters to "work involving routine, repetitive tasks and understanding and carrying out of simple instructions" and "only occasional superficial contact with co-workers." The RFC finding also excuses Walters from production quotas or strict time restraints because of his reduced tolerance for stress.
The ALJ's written decision appears to rest on an adverse credibility finding and two medical reports. He notes that for several years Walters had used his dead brother's social security number and, under that number, apparently earned wages from 2004 to 2007 that met the threshold for substantial gainful activity. Indeed, Walters admitted that for several years he had used the name and social security number of a brother who died in infancy. He explained that the men he suspects of killing his sister in a Texas drug deal also threatened to kill him after he reported his suspicions to the police. This fear, Walters said, prompted him to use his brother's identifiers to elude his sister's murderers. The ALJ stopped short of attributing to Walters the earnings recorded under his brother's social security number (which would have ended his disability claim at Step 1), but the ALJ did take the situation into account in assessing Walters's credibility. The ALJ also wrote that Walters's choice of timing to resume mental-health treatment (shortly after applying for benefits) suggests that he was motivated by monetary gain. The ALJ called Walters's representations to Quinco "inconsistent with the reports of ongoing work activity" and hobbies. Ultimately, the ALJ relied on Dr. Horton's opinion (that Walters's ability to concentrate is only moderately limited) as well as Dr. Karkut's opinion (that Walters suffers only mild depression and had achieved a GAF score of 70) but did not mention Dr. Evans's opinion (that Walters has poor concentration and poor social skills and had a GAF of just 45).
Walters's only argument in this court is that the ALJ erred by failing to explain the weight he gave to Dr. Evans's medical report. The Commissioner acknowledges that the ALJ's written decision fails to mention Dr. Evans or his observations, and the Commissioner also concedes that this omission constitutes error. But the Commissioner argues that Dr. Evans's report is not materially inconsistent with the rest of the medical evidence or the ALJ's decision. According to the Commissioner, substantial evidence supports the ALJ's ultimate decision that Walters is not disabled, and any error is harmless. We review the ALJ's decision as the final word from the Commissioner since the Appeals Council denied review, see Getch v. Astrue, 539 F.3d 473, 480 (7th Cir. 2008), and confine our review to the rationale offered by the ALJ, see SEC v. Chenery Corp., 318 U.S. 80, 93-95 (1943); Steele v. Barnhart, 290 F.3d 936, 941 (7th Cir. 2002).
We will uphold a denial of disability benefits as long as the decision is supported by substantial evidence and free of legal error. See 42 U.S.C. § 405(g); Nelms v. Astrue, 553 F.3d 1093, 1097 (7th Cir. 2009). ALJs need not mention every piece of evidence. Schmidt v. Barnhart, 395 F.3d 737, 744 (7th Cir. 2005); Diaz v. Chater, 55 F.3d 300, 308 (7th Cir. 1995). Nevertheless, a medical opinion from an examining consultative psychologist (like Dr. Evans) is not just another piece of evidence. See McKinzey v. Astrue, 641 F.3d 884, 891 (7th Cir. 2011). The SSA's own regulations provide that ALJs "may not ignore these opinions and must explain the weight given to these opinions in their decisions." SSR 96-6p, 1996 SSR LEXIS 3 (July 2, 1996); McKinzey, 641 F.3d at 891; see also 20 C.F.R. § 404.1527(f). These regulations are binding on ALJs. Lauer v. Apfel, 169 F.3d 489, 492 (7th Cir. 1999). Of course the regulations do not literally say that ALJs must explicitly mention every doctor's name and every detail in their reports. However, when there is reason to believe that an ALJ ignored important evidence — as when an ALJ fails to discuss material, conflicting evidence — error exists. See McKinzey, 641 F.3d at 891; Brindisi v. Barnhart, 315 F.3d 783, 786 (7th Cir. 2003). Otherwise, we cannot confidently assess the agency's rationale and afford the claimant meaningful review. See Scott v. Barnhart, 297 F.3d 589, 596 (7th Cir. 2002).
That is the situation here. The record is conflicting regarding Walters's concentration, persistence, and pace. Dr. Horton opined that Walters has only moderate problems with concentration, and Dr. Karkut says that Walters's immediate memory is intact. But Dr. Evans says tat Walters's immediate rote memory is low, likely related to poor concentration. The ALJ should have addressed the apparent conflict that exists among these different pieces of evidence. See Brindisi, 315 F.3d at 786; Scott, 297 F.3d at 596. It appears that the ALJ simply adopted Dr. Horton's language. It is true that Dr. Evans also wrote that Walters's attention span is below average but adequate to attend to a simple work routine. That assessment does not necessarily mean that Walters has the ability to concentrate on that routine for very long. The record is similarly conflicting regarding Walters's social skills. Dr. Horton observed only moderate limitations in social skills. But Dr. Evans observed "significant difficulties with paranoid thinking and poor social skills," which do not go away with medication. Again the ALJ did not discuss the difference, and instead preceeded to accept Dr. Horton's language. See Brindisi, 315 F.3d at 786; Scott, 297 F.3d at 596. Similarly, the ALJ cited GAF scores of 61 (from Quinco) and 70 (from Dr. Karkut) but ignored that Dr. Evans gave Walters a GAF of 45, only four months earlier than Dr. Karkut's evaluation. The evidence is inconsistent enough that the ALJ should discuss the conflict so that we can assess the agency's rationale and afford Walters meaningful review. See Scott, 297 F.3d at 596.
Moreover, nothing in the RFC leads us to believe that the ALJ accounted for Dr. Evans's assessment. The Commissioner argues that, by limiting Walters to routine, repetitive tasks with simple instructions, the ALJ accounted for Dr. Evans's finding that Walters suffers from poor concentration. Usually those terms will not account for poor concentration. See O'Connor-Spinner v. Astrue, 627 F.3d 614, 620 (7th Cir. 2010); Stewart v. Astrue, 561 F.3d 679, 684 (7th Cir. 2009); Craft v. Astrue, 539 F.3d 668, 677-78 (7th Cir. 2008); Kasarsky v. Barnhart, 335 F.3d 539, 544 (7th Cir. 2003). Limitations might be accounted for indirectly, but only if it is "manifest that the ALJ's alternative phrasing specifically excluded those tasks that someone with the claimant's limitations would be unable to perform." O'Connor-Spinner, 627 F.3d at 619. One example of this is when a claimant's limitations are entirely stress-related and the RFC limits the claimant to low-stress work. See Johansen v. Barnhart, 314 F.3d 283, 285, 288-89 (7th Cir. 2002). Here it is not manifest that limiting Walters to simple tasks is enough to enable him to work. As Dr. Evans noted, Walters "has significant difficulties with paranoid thinking and poor social skills even when taking his medication." Paranoia and poor social skills certainly hinder one's ability to concentrate on a task and those impairments do not necessarily diminish as a job gets easier. See O'Connor-Spinner, 627 F.3d at 619-20 (limiting claimant to repetitive tasks with simple instructions did not account for claimant's depression-related problems); Craft, 539 F.3d at 677-78 (limiting claimant to unskilled work did not account for claimant's mood swings). For similar reasons, the ALJ did not account for Dr. Evans's view that Walters has poor social skills just by limiting Walters just to superficial contact with coworkers. Perhaps someone with poor social skills should not have even superficial contact with workers. Because Walters's disability claim turns on that determination — the VE testified that no jobs exist for someone who cannot deal with coworker interaction — the ALJ should have provided a rationale to review. See Brindisi, 315 F.3d at 786-87; Scott, 297 F.3d at 596.
The Commissioner argues that "Walters waived his right to appellate review on the issue of whether the ALJ's hypothetical question to the vocational expert adequately accounted for his finding that Walters had moderate deficiencies in concentration" because he did not raise the issue to the district court. But Walters hasn't even raised that issue on appeal. Here he is not claiming independent error concerning the hypothetical question. He is simply — and correctly — citing O'Connor-Spinner to show that the phrases the ALJ used in his RFC finding and the terms in Dr. Evans's report are not functionally the same.
The Commissioner also attempts to discount the significance of Dr. Evans's assessment that Walters had a low GAF score of 45. The Commissioner maintains that the ALJ discussed the conflicting GAF scores during the hearing and again in his written decision. That contention is misleading. The ALJ discussed the conflicting high scores from Dr. Karkut and the staff at Quinco, but he did not mention the low one from Dr. Evans. In fact the ALJ's decision suggests consistency among Walters's GAF scores: The ALJ asserts that the "treatment records at Quinco consistently show Global Assessments of Function of 61," which he then compares to Dr. Karkut's similarly high GAF score of 70. The only speaker discussing the conflicting GAF scores at the page of the hearing transcript cited by the Commissioner is Walters's lawyer, not the ALJ. It is true that an ALJ does not need to base his conclusion entirely on GAF scores. See Howard v. Comm'r of Soc. Sec., 276 F.3d 235, 241 (6th Cir. 2002). Indeed, Dr. Evans's assessment, as the Commissioner argues, might be entitled to less weight than the others. However, the ALJ certainly thought GAF scores important, as he cited the high GAF scores from Quinco and Dr. Karkut. At this point we are not weighing the evidence, or even making sure that the ALJ's ultimate decision is supported by substantial evidence" rather, we are making sure that the ALJ considered the evidence. See McKinzey, 641 F.3d at 891-92; Spiva v. Astrue, 628 F.3d 346, 353 (7th Cir. 2010).
This last point requires more explanation. The Commissioner argues that the ALJ's decision is supported by substantial evidence, or "such relevant evidence as a reasonable mind might accept as adequate to support a conclusion." Richardson v. Perales, 402 U.S. 389, 401 (1971). Applying this standard would defeat the entire purpose of the doctrine of harmless error, which is to ensure that the first-line tribunal is not making serious mistakes or omissions. Spiva, 628 F.3d at 353. "When the decision of that tribunal on matters of fact is unreliable because of serious mistakes or omissions, the reviewing court must reverse unless satisfied that no reasonable trier of fact could have come to a different conclusion, in which event a remand would be pointless." Sarchet v. Chater, 78 F.3d 305, 309 (7th Cir. 1996). In this context, an error in failing to analyze and explain important evidence is not harmless simply because the ALJ could have addressed that evidence in a way that would survive substantial-evidence review. See Parker v. Astrue, 597 F.3d 920, 924-25 (7th Cir. 2010); Spiva, 628 F.3d at 353; Terry v. Astrue, 580 F.3d 471, 476 (7th Cir. 2009); Sarchet, 78 F.3d at 309.
We concluded that the error here is not harmless. This is not a case where the ALJ ignored the report of a nonexamining doctor who contradicted a treating physician after looking only at medical records. See McKinzey, 641 F.3d at 892 (finding harmless error); see also 20 C.F.R. § 404.1527(d)(1); Haynes v. Barnhart, 416 F.3d 621, 631 (7th Cir. 2005); Wildman v. Astrue, 596 F.3d 959, 967-68 (8th Cir. 2010). The Commissioner asserts that we can predict with great confidence that the ALJ would reach the same result on remand because Dr. Evans's lower GAF score is due to the fact that Walters was not regularly taking his medication at the time. See McKinzey, 641 F.3d at 892. That contention is both factually and legally inadequate. Factually, it is not clear that Walters was off his medication. Three weeks before Dr. Evans's examination, Nurse Kell at Quinco had restarted Walters on Risperdal (what is unclear is whether Walters was actually taking the drug or whether the medication had kicked in). Nothing in Dr. Evans's report suggests that Walters was not on his medication. Legally, that Walters may have been off his medication is not a significant fact in determining whether he is disabled; people with mental illness often struggle to stay on their drugs because of the adverse side effects. See Martinez v. Astrue, 630 F.3d 693, 697 (7th Cir. 2011); Spiva, 628 F.3d at 351. The evidence in this case is conflicting, and an administrative law judge who considers Dr. Evans's report might decide to award benefits. See Sarchet, 78 F.3d at 309. If accepted, a GAF of 45 suggests that Walters may be unable to work. See AM. PSYCHIATRIC ASS'N, supra, at 34. Thus, we reject the Commissioner's contention that a remand would be pointless.
The judgment of the district court is REVERSED, and the case is REMANDED to the agency for further proceedings.