dissenting.
My problem with the court’s resolution lies not in its choice of standard of review but in its reading of the evidence from which it concludes that “Paul Revere’s denial was reasonably supported by substantial evidence.”
I begin with the court’s reading of the report of Dr. Cook, a neurosurgeon. The essence of the doctor’s conclusions is the following:
It is my feeling that the decompressive surgery in his neck has been successful in relieving pressure of the spinal cord and allowing a partial though incomplete recovery of spinal cord function. Mr. Doyle now has a chronic pain problem ássociated with immobility in the neck. I do not think that the mild residual myelopathy impairs him from either the written or verbal description that I was given. I think that his disability centers around a chronic pain syndrome. The only measurable aspect of this is the associated immobility in the neck which is rather severe and detailed above. I do not think that any further testing is required to better elucidate the problem.
The various positions required of Mr. Doyle in order to perform his work, especially at the computer terminal, seem to exacerbate the pain that he has been having and, by his description, have made work intolerable to him. I think that any attempt at treatment would have to revolve around addressing these issues and seeing if there is any adjustment that can be made in his position of work to diminish the strain on the neck.
Appendix p. 127. My reading of these conclusions is that (1) Doyle has two sources of disability: residual myelopathy [or constriction of the spinal cord] and chronic pain syndrome; (2) the former does not impair him from the written or verbal description of his work; (3) but his disability “centers around a chronic pain syndrome,” of which a measurable aspect was his rather severe neck immobility; and (4) any treatment would have to address the pain problem and see if any adjustment in work position can diminish the strain on his neck.
The court, in my view, downplays the impact of pain by saying that the pain syndrome was acknowledged, rather than that Doyle’s disability “centers” around it. The doctor’s conclusion based on the residual myelopathy is not a conclusion as to his total disability. In addition, I read the doctor’s recommendations as to treatment as essential steps to the diminishment of pain, the center of his disability. I further see no suggestion in the record that the doctor, or in fact any other doctor, felt that Doyle was fabricating or exaggerating his pain. Dr. Cook’s conclusion that work restrictions would be in the realm of lifting, etc., was necessarily contingent on the amelioration of pain.
I also have a different understanding of what rehabilitation consultant Delahanty said in his first letter in January, 1991. Doyle, he acknowledged, had been “an outstanding employee.” He also wrote that Doyle “presentfed]” to him “an extremely disabled state with limited insight as to returning to competitive employment.” Delahanty passed no. present judgment on Doyle’s disability. Delahanty recognized that all his information had come from Doyle and that a Physical Capacities Evaluation was needed to establish “some base line functionary levels.”
*188In other words, I read Delahanty’s conclusion that “various job modifications are available ...” to depend on some ascertainment of what Doyle could really do. I see nothing to suggest that Delahanty “contemplated Doyle returning to work at Textron in a modified position” in the sense that he thought Doyle was at that time able to return to work.
The third piece of evidence relied on by the court was the Physical Capacity Evaluation of the New England Rehabilitation Center. After stating that, during the three-hour evaluation, Doyle “was limited in his activities due to poor endurance, pain, pain behaviors and decreased tolerance,” its concluding paragraph stated:
At this time, [patient] presents with a sedentary work capacity. To facilitate a gradual return to work, [patient] may benefit from a comprehensive pain program involving Psychology, Occupational Therapy, Physical Therapy and Biofeedback for symptom control, pain management strategies and reeducation as well as to increase functional status. If [patient] were to do well with the above, it is recommended that [patient] have a job site visit to alter or adjust the environment to facilitate proper body positioning thus decreasing .[patient’s] pain and allowing [patient] to return to work part time.
Appendix p. 168.
The court’s summary gave, in my view, short shrift to the report of pain, with no suggestion of contrivance on the part of Doyle, and again overlooked the conditional nature of the Center’s assessment of “a sedentary work capacity.” The “comprehensive pain program” was not merely to facilitate a “gradual return to work” in the sense of easing the return, but its success was a precondition. “If [patient] were to do well with the above,” then a job site visit to help body positioning would be the next step to “decreas[e] [patient’s] pain and allow[ ] [patient] to return to work part time.”
My reading of these ■ three reports supports the conclusion that Doyle’s spinal cord was not damaged such that it constituted a barrier to his being able to do part time, sedentary work of a nature suited to his background. But there was no discounting of the obstacle of pain as a barrier which would have to be surmounted. A comprehensive many-faceted pain program involving psychology, therapy, pain management strategies, and reeducation was recommended. Only if this had some success was even a return to part time work envisaged. This, in my opinion, is not substantial evidence supporting Paul Revere’s denial.
I would add two other pieces of evidence bearing on arbitrariness, or failure to act rationally. On March 27, 1991, the rehabilitation consultant Delahanty wrote Paul Revere about discussing with Doyle the possibility -of pursuing the Center’s recommendation of what he referred to as “therapeutic options, including work hardening.” He reported Doyle’s statement that his physician felt that he was “not a vocational rehabilitation candidate,” and that “it would be necessary for him to further discuss with his physician the possibility of any therapeutic options.... ”
Notwithstanding this statement, the consultant made no recommendation to follow up the Center’s suggestion or in any other way to address the pain problem. On the same day that Delahanty made his written report, Paul Revere wrote its letter denying Long Term Disability Benefits. There was no mention of discussing the details of a comprehensive pain program with Doyle’s treating physician. It referred to the reports of both Dr. Cook and the Center as “supporting] your ability to perform sedentary work,” with no mention of a precondition that the pain barrier must first be surmounted.
I make one final observation. I recognize, as does the court in its footnote 4, that Social Security Administration awards of disability benefits based on an inability to engage in substantial gainful employment are not of binding effect on disability insurers. But, in Doyle’s Social Security decision, the evidence of his “non-exertional limitations” (i.e., pain-related limitations) was found to trump what would otherwise be a capacity to engage in sedentary work. It was consistent with and supportive of my reading of the Cook and Center reports. Ignoring it in this case *189seems to me another indication of arbitrariness.
If indeed the court’s conclusion, affirming the denial of benefits, is commanded by law, an insurer may not only choose between doctors but may selectively read medical reports from the same doctors or evaluators,' selecting those parts which support its action and ignoring those which do not. If this can be equated with “substantial evidence,” I think that the administration of total disability policies is very substantially review-free.
I respectfully dissent.