Reset A A Font size: Print

United States Court of Appeals, Seventh Circuit.

KEITH A. ROBERTS, Plaintiff-Appellant, v. NANCY BERRYHILL, Acting Commissioner of Social Security, Defendant-Appellee.

No. 17-1533

Decided: January 22, 2018



In this Social Security appeal, the administrative law judge concluded that claimant Keith Roberts was not suffering from severe impairments as of September 30, 1982, and therefore was not disabled as of his date last insured. The district court agreed. We affirm, because substantial evidence in the record—or more accurately, the lack of evidence for Roberts's claim—supports the judge's decision.

Roberts (having been previously granted Supplemental Security Income) applied for Disability Insurance Benefits in March 2002 alleging that he became disabled in 1981 from coronary artery disease and post-traumatic stress disorder. Roberts served from 1968 to 1971 in the United States Navy, primarily as an enlisted airman at a non-combat facility in Italy. The basis of his claim of PTSD is an incident that he told doctors he witnessed in 1969: Airman Gary Holland, whom Roberts later described as a close friend, was killed when a plane he was working on collapsed. Roberts reported becoming angry and depressed after Holland's death. He submitted Navy records to the Social Security Administration that show that in December 1969, he was restrained in a straitjacket and given Thorazine (an anti-psychotic drug) when he became combative after a night of drinking. Then, in January 1970, Dr. James Bethel, a psychiatrist, evaluated Roberts and concluded that, although he showed signs of anger and immaturity, “no formal psychiatric diagnosis” was warranted. And twice in the summer of 1970, Roberts was prescribed Librium, a sedative that can treat anxiety.

After he left the Navy, Roberts worked intermittently, but stopped working by 1982. His earnings record shows (and Robert does not contest) that he was last insured for benefits purposes on September 30, 1982. See 42 U.S.C. § 423(a)(1)(A), (c). Yet besides his sporadic treatment for anxiety while in the Navy, the record contains no significant medical or psychiatric treatment notes from before that date. The only other record from before Roberts's date last insured is a treatment note from February 1982, explaining that Roberts was having urinary problems, which are not relevant now.

The record does show, however, that after September 1982, Roberts's mental and physical health declined. In January 1987, he had a heart attack and pulmonary embolism. In March 1991, he discussed Holland's death with Dr. James Hastings, whom he saw at a Department of Veterans Affairs medical center for a psychiatric exam. Dr. Hastings did not diagnose Roberts with PTSD but said anger was Roberts's “major problem” and diagnosed dysthymia and mixed personality disorder. In November 1992, Roberts saw Dr. K. Sherry, who agreed and added that Roberts suffered from alcohol abuse disorder. Dr. Sherry also noted that Roberts had been taking Prozac, which had helped his anger and depression. Also in November 1992, Dr. Jonathan Thomas of Bay Psychiatric Clinic noted that Roberts had sought treatment there since November 1989 for major depression, but that Prozac was helping his anger and anxiety. In March 1993, Roberts again saw Dr. Hastings, who again diagnosed dysthymia and mixed personality disorder without discussing the possibility of PTSD. That month, the Department of Veterans Affairs granted Roberts disability benefits retroactively to November 1990 based in part on his “dysthymia with history of major depressive disorder and mixed personality disorder.”

The first medical diagnosis of post-traumatic stress disorder in the record is from March 1998, when Dr. Michael Daly examined Roberts at a Veterans Affairs medical center and concluded that he met the criteria for the disorder. Dr. Daly noted that in 1995 Roberts had received a score in the “positive range” for PTSD on a diagnostic test, which is not in the record. The Department of Veterans Affairs then granted Roberts additional disability benefits. In April 1999, Dr. Michael Marcy diagnosed PTSD, and the Department increased Roberts's benefits again. Since then, some clinicians, including an expert Roberts consulted for his Disability Insurance Benefits claim (Dr. Donald Derozier), a Veterans Affairs psychiatrist (Dr. Chandra Bommakanti), and a consulting agency doctor (Dr. Timothy Lynch) have opined that there may be support for a PTSD diagnosis dating back to Roberts's time in the Navy. But Dr. Hastings, who saw Roberts in 1991 and 1993, said in 2003 that he did not believe Roberts met the criteria for the disorder at their first two meetings, much less before 1982.

Roberts's application for Disability Insurance Benefits has been winding its way through the Social Security Administration and the courts for over a decade. He has been both granted and denied benefits in administrative decisions that, but for the most recent one, have all been vacated and remanded for further proceedings. The most recent decision—the one at issue in this appeal—was entered in 2014 after a hearing before an administrative law judge at which Roberts, his wife, and a vocational expert testified. The judge found that Roberts showed symptoms of PTSD and anxiety (but not any heart conditions) before his date last insured of September 30, 1982. But, the judge concluded, neither condition was severe before that date. Thus Roberts was ineligible for Disability Insurance Benefits. The Appeals Council denied review, so the judge's decision is the final word of the Commissioner. See Engstrand v. Colvin, 788 F.3d 655, 660 (7th Cir. 2015). The district court affirmed the decision.

Substantial evidence supports the administrative law judge's conclusion that Roberts was not severely impaired by any of his conditions before September 30, 1982. If a claimant is not disabled at the time he last has social security insurance coverage, he is not eligible for Disability Insurance Benefits even if he later becomes disabled. Martinez v. Astrue, 630 F.3d 693, 699 (7th Cir. 2011). As for Roberts's heart condition—about which even Roberts does not say much on appeal—the judge correctly pointed out that no evidence showed that Roberts had any problems before his 1987 heart attack. And regarding Roberts's mental health, it is unclear whether Roberts met the criteria for post-traumatic stress disorder before 1990. The judge correctly noted that no evidence showed Roberts suffered from any functional limitations from PTSD or anxiety before his date last insured. The only clinician to examine Roberts before then—Dr. Bethel in 1970—concluded that no formal diagnosis was warranted and that Roberts could successfully complete his enlistment.1 The sources who opined that Roberts may have had PTSD earlier than the 1990s—Dr. Derozier, Dr. Bommakanti, and Dr. Lynch—did not evaluate him until much later, and their opinions do not address or even imply the presence of functional limitations as far back as September 1982. The judge's conclusion that Roberts's conditions were not severe enough to cause disabling limitations as of his date last insured was therefore reasonable. See Martinez, 630 F.3d at 699.

Roberts contends, however, that the administrative law judge gave short shrift to the opinions of clinicians who treated him. See 20 C.F.R. § 404.1527(c)(2). But the only opinion he refers to that is in the administrative record is from a psychologist, Dr. Derozier, who examined him for purposes of helping him obtain benefits, making him a “nontreating” source. See § 404.1527(a)(2) (“treating source” excludes those without ongoing treatment relationship and those consulted “solely” for support for disability claim); Simila v. Astrue, 573 F.3d 503, 514 (7th Cir. 2009) (same). More important, our review of the record reveals that the judge thoroughly evaluated all the opinions from treating and nontreating sources alike and carefully explained his reasons for accepting or rejecting them. See 20 C.F.R. § 404.1527(c)(2) (explaining how to evaluate opinions by treating sources). The judge did not clearly err in evaluating the medical opinions, none of which describe functional limitations earlier than September 1982 anyway. See Stepp v. Colvin, 795 F.3d 711, 718–19 (7th Cir. 2015).

Roberts responds, however, that the administrative law judge impermissibly “played doctor.” See Goins v. Colvin, 764 F.3d 677, 680 (7th Cir. 2014) (independently analyzing medical evidence is “a clear no-no”). First, he says, the judge should have consulted a medical expert about Roberts's PTSD diagnosis. This contention is meritless. Consulting a new expert, who could have done little more than evaluate a cold record more than thirty years after Roberts's date last insured, was unlikely to yield any information not already contemplated by the other clinicians who had reviewed Roberts's medical records. We therefore will not disturb the judge's discretionary decision not to consult an additional expert to interpret the medical records. See Nelms v. Astrue, 553 F.3d 1093, 1098 (7th Cir. 2009) (“This court generally upholds the reasoned judgment of the Commissioner on how much evidence to gather, even when the claimant lacks representation.”). Roberts also contends that the judge assumed without medical evidence that one cannot suffer from PTSD without experiencing combat. But Roberts misreads the judge's opinion. The judge merely noted that Roberts had not seen combat, most likely to provide background and to clarify what stressors Roberts identified as causes of his PTSD.

Roberts next argues that the administrative law judge should have followed earlier administrative decisions that were favorable to him. He points to both the decisions from the Department of Veterans Affairs granting him disability benefits and earlier opinions by a Social Security administrative law judge granting him Disability Insurance Benefits (the latter of which have since been vacated). Roberts is wrong on both counts. First, administrative law judges must consider, but not necessarily give controlling weight to, decisions by other agencies such as the Department of Veterans Affairs. 20 C.F.R. § 404.1504; Bird v. Berryhill, 847 F.3d 911, 913 (7th Cir. 2017). The Department of Veterans Affairs found Roberts disabled beginning, at the earliest, in 1990—eight years after his date last insured. The administrative law judge reasonably discounted these opinions. See Martinez, 630 F.3d at 699. Second, the judge was correct not to rely on earlier decisions by the Social Security Administration about Roberts's application for Disability Insurance Benefits. Those decisions were properly vacated and remanded by the Appeals Council, and thus could not be binding at a later point in the case. See Aguiniga v. Colvin, 833 F.3d 896, 900–01 (8th Cir. 2016) (vacated administrative decisions have no binding effect on later decisions).

Roberts's next argument is that these proceedings have been “tainted” by “allegations of fraud” against him, which he says caused the administrative law judge to be biased against him. He refers to his Veterans Affairs file, which is not in the record, but which he says contains requests from the Department of Veterans Affairs for Social Security records that are stamped “Please Expedite Fraud Case.” We are aware, however, that what Roberts refers to as “allegations” actually resulted in his conviction in 2005 for fraudulently obtaining disability benefits from the Department, as described in United States v. Roberts, 534 F.3d 560 (7th Cir. 2008) (affirming 48-month prison sentence). The district court correctly recognized that this conviction could not be used as a basis for affirming the administrative law judge's decision in this case because it does not appear in the administrative record. See 42 U.S.C. § 405(g). We therefore will not address Roberts's objection further, except to clarify one issue: If, as Roberts seems to believe, the judge became aware of the fraud “allegations” during the pendency of his application, the judge would have been obligated to address them. See 42 U.S.C. § 405(u)(1)(A) (“The Commissioner of Social Security shall immediately redetermine the entitlement of individuals to monthly insurance benefits ․ if there is reason to believe that fraud or similar fault was involved in the application. ․ ”). That the judge did not speaks more to Roberts's luck than to any bias on the judge's part.

Still, Roberts contends that the district court should have remanded the case for consideration of other new evidence, including a 2015 psychiatric evaluation, Navy records, a competence evaluation from his 2005 federal criminal case, a 2010 opinion from the U.S. Court of Appeals for Veterans Claims, and the transcript of an agency hearing from 1994. A remand may be required “upon a showing that there is new evidence which is material and that there is good cause for the failure to incorporate such evidence into the record in a prior proceeding.” 42 U.S.C. § 405(g). All but the 2015 opinion were available to Roberts before his 2013 hearing, so he should have submitted them then. And the 2015 opinion of Dr. Robert Marcellino is not relevant or material. Dr. Marcellino merely opined, as other clinicians have, that Roberts now suffers from, or shows symptoms of, PTSD and that it might have begun after his time in the Navy. This opinion sheds no additional light on the key issue: whether Roberts had disabling limitations before September 30, 1982. It likely would not change the outcome in the event of a remand. See id.; Simila, 573 F.3d at 522.

We have considered Roberts's other arguments and none has merit. Accordingly, we AFFIRM the district court's judgment upholding the denial of Roberts's application for Disability Insurance Benefits.


1.   We have not seen it noted in the record, and Roberts has not pointed out, that PTSD did not exist as a diagnosable condition in 1970. It did not enter the American Psychiatric Association's Diagnostic and Statistical Manual of Mental Disorders until the third volume, published in 1980. Even then, the diagnosis was “controversial.” See Matthew J. Friedman, PTSD History and Overview, U.S. DEPARTMENT OF VETERANS AFFAIRS (Feb. 23, 2016), We do not think this point of interest affects the result here, though, because the DSM-III was published before Roberts's date last insured and because of the absence of evidence that Roberts had any functional limitations as of 1982.