FILED
                                                  United States Court of Appeals
                     UNITED STATES COURT OF APPEALS       Tenth Circuit

                            FOR THE TENTH CIRCUIT                       October 17, 2012
                                                                      Elisabeth A. Shumaker
                                                                          Clerk of Court
GLORIA BORGSMILLER,

             Plaintiff-Appellant,

v.                                                          No. 11-5160
                                                  (D.C. No. 4:10-CV-00481-TLW)
MICHAEL J. ASTRUE, Commissioner                             (N.D. Okla.)
of the Social Security Administration,

             Defendant-Appellee.


                            ORDER AND JUDGMENT *           0F




Before HARTZ, ANDERSON, and EBEL, Circuit Judges.


      Gloria Borgsmiller appeals from an order of the district court affirming the

Commissioner’s decision denying her application for disability benefits under the

Social Security Act. We exercise jurisdiction under 28 U.S.C. § 1291 and 42 U.S.C.

§ 405(g), and reverse and remand for further proceedings.




*
      After examining the briefs and appellate record, this panel has determined
unanimously to grant the parties’ request for a decision on the briefs without oral
argument. See Fed. R. App. P. 34(f); 10th Cir. R. 34.1(G). The case is therefore
ordered submitted without oral argument. This order and judgment is not binding
precedent, except under the doctrines of law of the case, res judicata, and collateral
estoppel. It may be cited, however, for its persuasive value consistent with
Fed. R. App. P. 32.1 and 10th Cir. R. 32.1.
                                      I. Background

        In 1999, Ms. Borgsmiller had back surgery due to a herniated disc. Although

the surgery was initially successful in relieving her back pain, she began to

experience pain again a few years later. In November 2001, she was terminated from

her employment as an office manager and has not engaged in substantial work since

that time. She applied for disability benefits in September 2006, alleging disability

beginning on November 30, 2001, due to a back injury, arthritis, hypothyroidism, and

high blood pressure. She claimed that due to these conditions she could not sit,

stand, or walk for any length of time and was bedridden for three to four weeks at a

time.

        The facts of Ms. Borgsmiller’s medical history concerning her back pain are

known to the parties and so we only summarize them here. 1 After her back surgery,
                                                             1F




Ms. Borgsmiller first sought treatment for back pain in 2002, complaining of pain in

her back that radiated down her left leg. She was diagnosed with lower back pain

with radiculopathy, to be treated with pain medication and a muscle relaxant. The

medical evidence submitted does not document further treatment for back pain until

2006.

        In March 2006, she presented to Timothy Sanford, D.O., who became her

treating physician, and reported that she had “flares of [back] pain” and, when severe,
1
       Because Ms. Borgsmiller does not claim error on appeal with respect to her
other conditions for her alleged disability, i.e., arthritis, hypothyroidism, and high
blood pressure, we focus, as she does, on her back pain.


                                          -2-
she hurt for three to eight weeks and was bedbound. Admin. R. at 166. He assessed

low back pain, a herniated lumbar disc, and lumbar radiculopathy, and prescribed

pain medication and a muscle relaxant. She periodically treated with Dr. Sanford

between 2006 and 2008, reporting intermittent flares of back pain that she claimed at

times left her bedbound for weeks or months at a time. Treatment notes from 2006

reflect that Dr. Sanford opined that Ms. Borgsmiller’s herniated disc and lumbar

radiculopathy interfered with her activities of daily living. He further opined that she

was unable to work due to periods of exacerbation of back pain which left her

bedbound. Ms. Borgsmiller continued treatment with pain medication and muscle

relaxants, prescribed by Dr. Sanford, but did not otherwise have other treatment for

her back pain. 2 A consultative examination done in November 2006 confirmed that
              2F




Ms. Borgsmiller suffered from low back pain, leg pain, had a limited range of motion

of her spine, and ambulated with an unstable gait at a slow speed with the use of a

cane. An agency physician reviewed Ms. Borgsmiller’s records and concluded she

had a residual functional capacity (RFC) to perform light work. In March 2007, the

evidence was again reviewed by the agency and her light RFC was affirmed.

      The Commissioner denied Ms. Borgsmiller’s application initially and on

reconsideration. After a de novo hearing before an administrative law judge (ALJ),

the ALJ issued her decision in August 2008, finding Ms. Borgsmiller not disabled at
2
       Treatment notes reflect Dr. Sanford referred Ms. Borgsmiller to a neurologist
and recommended she undergo an MRI to further examine her lumbar radiculopathy,
but she declined due to cost issues and the lack of health insurance.


                                          -3-
step four of the controlling five-step sequential analysis. See Lax v. Astrue, 

, 1084 (10th Cir. 2007) (explaining five-step process for evaluating claims for

disability benefits). 3 The ALJ confirmed that Ms. Borgsmiller had not worked during
                   3F




the period from her alleged onset of disability, November 30, 2001, through her date

last insured, December 31, 2006. She found that Ms. Borgsmiller had a severe

impairment: degenerative disc disease of the lumbar spine. But she held that this

impairment did not meet or equal the listing for presumptive disability for disorders

of the spine, 20 C.F.R. Pt. 404, Subpt. P, App. 1, § 1.04.

      According to the ALJ, this impairment left Ms. Borgsmiller with an RFC to

perform sedentary work. Relying on the Dictionary of Occupational Titles (DOT)

and associated inquiries to the vocational expert (VE) who testified at the hearing,

the ALJ concluded that Ms. Borgsmiller could still perform her past work as a service

office manager and thus was not disabled.

      Ms. Borgsmiller’s request for review was denied by the Appeals Council,

making the ALJ’s decision the Commissioner’s final decision for purposes of judicial

review. Krauser v. Astrue, 

, 1327 (10th Cir. 2011). She then filed an

action in the district court and it affirmed the Commissioner’s denial of benefits.


3
       Although not pertinent here, the agency recently revised its rules regarding the
five-step sequential analysis such that an ALJ may proceed to step five if she
determines that there is insufficient evidence about a claimant’s past relevant work to
make a finding at step four of the analysis. See Expedited Vocational Assessment
Under the Sequential Evaluation Process, 77 Fed. Reg. 43492-01 (July 25, 2012)
(codified at 20 C.F.R. § 404.1520(h)).


                                          -4-
Ms. Borgsmiller now appeals.

                                      II. Discussion

      “In reviewing the [Commissioner’s] decision, we neither reweigh the evidence

nor substitute our judgment for that of the agency.” Branum v. Barnhart, 

, 1270 (10th Cir. 2004) (internal quotation marks omitted). Rather, “[w]e

review the Commissioner’s decision to determine whether the factual findings are

supported by substantial evidence in the record and whether the correct legal

standards were applied.” Watkins v. Barnhart, 

, 1299 (10th Cir. 2003).

“Substantial evidence is such relevant evidence as a reasonable mind might accept as

adequate to support a conclusion. It requires more than a scintilla, but less than a

preponderance.” 

(citation omitted) (internal quotation marks

omitted).

      Ms. Borgsmiller raises three challenges to the Commissioner’s decision: the

ALJ (1) should have recontacted Ms. Borgsmiller’s treating physician to clarify the

record regarding her flares of back pain; (2) erred in her credibility determination;

and (3) did not perform a proper step-four analysis. She also asserts claims of error

by the magistrate judge. “[W]e independently determine whether the ALJ’s decision

is free from legal error and supported by substantial evidence.” Wall v. Astrue,

, 1052 (10th Cir. 2009) (internal quotation marks omitted). Therefore,

although we have considered these arguments about the magistrate judge in the




                                          -5-
context of evaluating her claims that the ALJ erred, we do not expressly address her

claims of error by the magistrate judge.

   A. Recontacting the Treating Physician

      Based on the record before her, the ALJ gave “great weight to Dr. Sanford’s

opinion that [Ms. Borgsmiller was] disabled during her flares” but concluded that her

flares of pain “[did] not occur often enough to prevent her from engaging in

substantial gainful activity.” Admin. R. at 16. On appeal, Ms. Borgsmiller

challenges the evidentiary support for this conclusion, in light of the weight given to

Dr. Sanford’s opinion, and argues that the ALJ should have recontacted Dr. Sanford

because the frequency or number of flares in the record is unclear. Relatedly, she

also claims the ALJ impermissibly inferred that Ms. Borgsmiller was not disabled

between her flares.

      Ms. Borgsmiller bears the burden of proving her disability and must do so by

furnishing medical and other evidence of the existence of her disability. See 

. On the other hand, the ALJ has a basic obligation to “ensur[e] that

an adequate record is developed during the disability hearing consistent with the

issues raised.” 

omitted). As applicable at the time that

Ms. Borgsmiller’s claim for disability benefits was adjudicated, an ALJ was required

to recontact a treating source when the evidence received from a treating source was

deemed inadequate for the ALJ to determine whether a claimant was disabled.

White v. Barnhart, 

, 908 (10th Cir. 2001); see also Robinson v.


                                           -6-
Barnhart, 

, 1084 (10th Cir. 2004) (noting ALJ’s duty to seek additional

evidence or clarification from treating source when evidence contains conflict or

ambiguity, does not contain all the necessary information, or does not appear to be

based on medically acceptable clinical and laboratory diagnostic techniques). 4 The
                                                                                 4F




ALJ’s duty in that regard was not dependent on any further action by the claimant

once the evidence was deemed inadequate. 

.

          Ms. Borgsmiller does not point to evidence that states or suggests that the ALJ

deemed the evidence from Dr. Sanford inadequate or ambiguous, and we find none.

But Ms. Borgsmiller suggests the evidence was, in fact, inadequate or ambiguous

because “the record is unclear as to the number of [her] flares” and “[t]he frequency

of [her] disabling flares is open to question.” Aplt. Br. at 16-17. Based on our

review, it does not appear, however, that further contact with Dr. Sanford would have

provided the ALJ more clarity. Other than one medical visit with Dr. Sanford in

March 2006, Dr. Sanford did not otherwise diagnose a flare or examine

Ms. Borgsmiller when a flare was occurring. Rather, Dr. Sanford’s medical notes

reflect only Ms. Borgsmiller’s self-reporting of her flares and the duration of those

flares.


4
       Effective March 26, 2012, the regulations governing an ALJ’s “duty” to
recontact a medical source have changed. Under the current regulations, in
evaluating insufficient evidence to determine disability, an ALJ “may recontact [a]
treating physician, psychologist, or other medical source” but may instead seek
further evidence from another source, including the claimant herself. See 20 C.F.R.
§ 404.1520b.


                                            -7-
      Accordingly, if the number and/or frequency of Ms. Borgsmiller’s flares are

unclear, Dr. Sanford is of no assistance in clarifying the matter. And

Ms. Borgsmiller has not argued that there are existing reports that Dr. Sanford could

have provided the ALJ that would have assisted her in ascertaining the basis for his

opinions. See e.g., 

-72 (concluding that the ALJ

satisfactorily developed the record when the claimant’s “counsel did not indicate or

suggest to the ALJ that any medical records were missing from the administrative

record, nor did counsel ask for the ALJ’s assistance in obtaining any additional

medical records”). We conclude that the ALJ was not obligated to recontact

Dr. Sanford to clarify any alleged ambiguity regarding the number and/or frequency

of Ms. Borgsmiller’s flares.

      Ms. Borgsmiller also challenges the ALJ’s conclusion that “Dr. Sanford’s

opining that claimant was disabled during the flares implies that she was not disabled

between the flares.” Admin. R. at 16; see also Aplt. Br. at 16, 18. To the extent

Ms. Borgsmiller contends the ALJ should have recontacted Dr. Sanford instead of

inferring that Ms. Borgsmiller was not disabled in between flares, we disagree.

Dr. Sanford’s opinion that Ms. Borgsmiller is unable to work during her flares of

severe pain is not a medical opinion, 20 C.F.R. § 404.1527(e), but an opinion on the

ultimate issue of disability, which is a matter reserved to the Commissioner. See

Castellano v. Sec’y of Health & Human Servs., 

, 1029 (10th Cir. 1994).

As such, the ALJ was not required to recontact Dr. Sanford on this ground either.


                                         -8-
   B. Credibility Determination

      Ms. Borgsmiller next challenges the ALJ’s credibility determination, with

respect to her complaints of pain. “Credibility determinations are peculiarly the

province of the finder of fact, and we will not upset such determinations when

supported by substantial evidence.” Hackett v. Barnhart, 

,

1173 (10th Cir. 2005) (internal quotation marks omitted). “However, findings as to

credibility should be closely and affirmatively linked to substantial evidence and not

just a conclusion in the guise of findings.” 

quotation

marks omitted).

      “To be disabling, pain must be so severe, by itself or in conjunction with other

impairments, as to preclude any substantial gainful employment.” Talley v. Sullivan,

, 587 (10th Cir. 1990) (internal quotation marks omitted). In assessing a

claimant’s complaints of disabling pain, an ALJ must consider both objective and

subjective evidence. See Kepler v. Chater, 

, 390 (10th Cir. 1995).

Because they are subjective, a claimant’s statements regarding her pain “can be

evaluated only on the basis of credibility.” Thompson v. Sullivan, 

,

1488-89 (10th Cir. 1993). As such, “[subjective] statements regarding the intensity

and persistence of the pain must be consistent with the medical findings and signs.”

Gossett v. Bowen, 

, 806 (10th Cir. 1988).

      In evaluating the credibility of a claimant’s subjective statements regarding

pain, an ALJ should consider an “individual’s daily activities;” the “location,


                                          -9-
duration, frequency, and intensity of the individual’s pain;” the “dosage and

effectiveness of medications taken to alleviate pain;” and “[a]ny other factors

concerning the individual’s functional limitations and restrictions due to pain or other

symptoms.” Hamlin v. Barnhart, 

, 1220 (10th Cir. 2004) (alteration in

original); see also 

(listing other factors such as

frequency of medical contacts).

      The ALJ determined Ms. Borgsmiller had a RFC at the sedentary level. In

doing so, the ALJ “d[id] not discount all of [Ms. Borgsmiller’s] complaints” of pain,

but found that her treating physicians “did not place any functional restrictions on her

activities that would preclude sedentary work activity with the previously-mentioned

restrictions.” Admin. R. at 14. After discussing portions of both the objective and

subjective evidence, the ALJ concluded that Ms. Borgsmiller’s “statements

concerning the intensity, persistence and limiting effects of [her] symptoms are not

credible to the extent they are inconsistent with [her] residual functional capacity

assessment.” 

reaching this credibility finding, the ALJ stated that Ms. Borgsmiller’s

flares of pain were “episodic . . . [and] occurring with decreasing frequency.” 

support, she pointed to medical evidence indicating a reported flare in

October 2007, one in February 2008, and none between February 2008 and the time

of the hearing in July 2008. And she stated that Dr. Sanford’s opinion that

Ms. Borgsmiller was disabled during a flare implied that she was not disabled


                                         - 10 -
between the flares. She discounted Ms. Borgsmiller’s allegations of disabling pain

from arthritis and carpal tunnel syndrome by pointing to medical evidence showing a

grip strength rated 5/5, and observed that while Ms. Borgsmiller used a cane, none

was prescribed by a physician. Finally, she stated that Ms. Borgsmiller remained

able to do chores, wash dishes, and prepare meals.

       On appeal, Ms. Borgsmiller contends that her “claim for disability does not

hinge on limitations with her hands . . . [or] on her use of a cane,” but, instead,

“revolves around her pain, and the limiting effect of her flares.” Aplt. Br. at 21. She

argues then that the ALJ’s credibility determination rested in large part on the

decreasing frequency of Ms. Borgsmiller’s flares and her daily activities, neither of

which substantially support a finding that Ms. Borgsmiller’s complaints are

incredible.

       As a preliminary matter, because the ALJ credited Dr. Sanford’s opinion that

Ms. Borgsmiller is unable to work during flares of severe pain, the ALJ consequently

must have found credible Ms. Borgsmiller’s subjective statements concerning the

intensity and limiting effects of the pain she experienced during a flare of severe

pain. Accordingly, although the ALJ did not specify so in her decision, we construe

her adverse credibility determination to correspond only with Ms. Borgsmiller’s

subjective statements concerning pain that she experienced during periods of

non-flare activity.




                                          - 11 -
      Reviewing the ALJ’s credibility determination in this manner, we are not

convinced that the ALJ’s rejection of Ms. Borgsmiller’s complaints of pain was

supported by substantial evidence. There are ample reports in the record of

Ms. Borgsmiller’s pain and prescriptions for pain medication. See 

(holding that evidence relied on by ALJ was insufficient to undermine

pain allegations where medical records were “replete with [claimant’s] reports of

pain and of prescriptions”). Yet, as Ms. Borgsmiller correctly asserts, the only

justifications given by the ALJ to discount Ms. Borgsmiller’s complaints were the

decreasing frequency of her flares and an alleged ability to perform daily activities.

      Implicit in the ALJ’s reliance on the alleged decreasing frequency of

Ms. Borgsmiller’s flares is a determination that the periods of non-flare activity

during which Ms. Borgsmiller could work were increasing to the point that the flares

had no significant impact on her overall ability to work. But the relevant period

within which to assess the disabling impact of Ms. Borgsmiller’s flares was before

December 31, 2006, when Ms. Borgsmiller was last insured for benefits. That

Ms. Borgsmiller’s flares in 2007 and 2008 may have been so infrequent as to have no

substantial impact on her ability to work does not necessarily support an inference

that their frequency before 2007 had a similarly insubstantial impact.

      Ms. Borgsmiller testified that in the beginning, she experienced flares two or

three times a year, the longest of which lasted three months. She further testified that

she had been able to do better in the last few years by staying at home and by not


                                         - 12 -
walking down stairs or getting into or out of a car. The ALJ did not assess the impact

of Ms. Borgsmiller’s flares before 2007 or evaluate the credibility of her allegations

that she could not work due to her pain during that earlier period. 5 In this, the ALJ
                                                                      5F




erred.

          In further support of her credibility determination, the ALJ relied on

Ms. Borgsmiller’s ability on “good days” to do chores, wash dishes, and prepare

meals. 6 Ms. Borgsmiller testified that she is able to do household chores
         6F




“sometimes,” Admin. R. at 32, and clarified as follows:

                [t]here are days that I can get up and I can . . . feel better
                and I can do some things. There are other times where I’m
                either sitting in my recliner or I’m in bed and there are
                other times I do not get out of bed for weeks at a time.

that she is unable to drive a vehicle, and “sometimes” has

difficulty getting dressed, showered, or even retrieving food from the refrigerator.

Although the ALJ did not err in considering Ms. Borgsmiller’s activities of daily


5
        Even if the decreasing frequency of Ms. Borgsmiller’s flares in 2007 and 2008
justified a determination that she was not disabled at the time of the hearing (a matter
we need not decide), it did not rule out the possibility that she was disabled for some
period before 2007 and thus entitled to a closed period of benefits.
6
       It appears that the ALJ equated Ms. Borgsmiller’s “good days” with non-flare
days and “bad days” with flare days. Such an assessment is problematic, however,
because Ms. Borgsmiller testified that she is in pain “most of the time,” Admin. R.
at 28, and has leg pain “all the time,” 

Ms. Borgsmiller may
experience pain even on “good days.” The question is how severe or limiting that
pain may be to determine whether Ms. Borgsmiller may engage in substantial gainful
activity. Accordingly, the ALJ’s assessment is not necessarily accurate.



                                            - 13 -
living, she “may not rely on minimal daily activities as substantial evidence that a

claimant does not suffer disabling pain.” 

. Indeed, the

“sporadic performance [of household tasks or work] does not establish that a person

is capable of engaging in substantial gainful activity.” 

(internal quotation marks omitted).

      Accordingly, although the ALJ tied her credibility finding to some specific

evidence in the record, Qualls v. Apfel, 

, 1372 (10th Cir. 2000), we

conclude none of the evidence relied upon, neither the decreasing frequency of flares

nor Ms. Borgsmiller’s alleged ability to do some daily activities, provides substantial

evidence that Ms. Borgsmiller’s complaints of pain were incredible. 7 See Hamlin,
                                                                       

. Therefore, we reverse and remand, directing the ALJ to

properly evaluate the evidence with respect to Ms. Borgsmiller’s credibility. We

express no opinion, however, as to the weight the Commissioner should give

Ms. Borgsmiller’s complaints of pain.


7
      In further support of her credibility determination, the ALJ also stated that
“Dr. Sanford’s opining that the claimant was disabled during the flares implies that
she was not disabled between the flares,” Admin. R. at 16. Even assuming,
arguendo, that the record supports a determination that Ms. Borgsmiller can work
during non-flare periods, as we discuss infra, the pertinent issue in determining
whether Ms. Borgsmiller is disabled within the meaning of the Act is whether she can
work on a regular and continuing basis. Accordingly, there must be substantial
evidence that Ms. Borgsmiller can engage in substantial gainful activity on a regular
and continuing basis despite her flares of severe pain during which, the ALJ
acknowledged, she could not work.




                                          - 14 -
   C. Step Four Determination

      Lastly, Ms. Borgsmiller contends that the ALJ erred in her RFC determination

at phase one of step four by failing to include any limitations from her flares in the

RFC assessment and in the inquiries to the VE at the hearing. She also contends this

improper RFC determination “corrupted [the ALJ’s] findings at the third phase”

regarding Ms. Borgsmiller’s ability to perform her past relevant work (PRW). Aplt.

Br. at 24. We agree.

             Step four of the sequential analysis . . . is comprised of
             three phases. In the first phase, the ALJ must evaluate a
             claimant’s physical and mental residual functional capacity
             (RFC), . . . and in the second phase, he must determine the
             physical and mental demands of the claimant’s past
             relevant work. . . . In the final phase, the ALJ determines
             whether the claimant has the ability to meet the job
             demands found in phase two despite the mental and/or
             physical limitations found in phase one. . . . At each of
             these phases, the ALJ must make specific findings.

Bowman v. Astrue, 

, 1272 (10th Cir. 2008) (alterations in original)

(citations omitted) (internal quotation marks omitted). Regarding the first phase, an

ALJ should “assess the nature and extent of [the claimant’s] physical limitations and

then determine [the claimant’s] residual functional capacity for work activity on a

regular and continuing basis.” Winfrey v. Chater, 

, 1023 (10th Cir.

1996) (alterations in original) (internal quotation marks omitted). RFC represents

“the most [that the claimant] can still do despite [her] limitations.” 20 C.F.R.

§ 404.1545(a)(1). To determine a claimant’s RFC, the ALJ must consider any

medical opinions about what the claimant can still do as well as the claimant’s own

                                         - 15 -
testimony concerning her limitations, including limitations resulting from pain. 8 

      The ALJ concluded Ms. Borgsmiller had the RFC to lift/carry ten pounds,

stand/walk for two hours during an eight-hour workday, only occasionally stoop, and

frequently climb, balance, kneel, crouch, and crawl. In reaching this conclusion, she

stated that Ms. Borgsmiller’s “treating physicians did not place any functional

restrictions on her activities that would preclude sedentary work activity with the

previously-mentioned restrictions.” Admin. R. at 14.

      With respect to Ms. Borgsmiller’s flares of severe pain, the ALJ gave “great

weight” to Dr. Sanford’s opinion that Ms. Borgsmiller is unable to work during her

flares. 

the ALJ concluded, based on the medical records showing

episodic reports of flares and decreasing flares, that Ms. Borgsmiller’s flares did not

occur often enough to prevent her from engaging in substantial gainful activity. See

20 C.F.R. § 404.1505(a) (noting that to be disabled, claimant must have impairment

that lasts not less than twelve months and causes an inability to engage in any

substantial gainful activity); see also Washington v. Shalala, 

,

1442 (10th Cir. 1994) (noting that finding that claimant is able to engage in



8
      In light of our determination regarding the ALJ’s credibility determination,
because the ALJ’s credibility and RFC determinations are inherently intertwined, any
new determination on credibility will necessarily require reassessment of
Ms. Borgsmiller’s RFC. See Poppa v. Astrue, 

, 1171 (10th Cir. 2009).



                                         - 16 -
substantial gainful activity requires determination that claimant can hold whatever

job she finds for a significant period of time).

       On appeal, Ms. Borgsmiller first claims that the ALJ’s RFC assessment is

improper because it did not include any limitations from her flares. We agree.

Although Dr. Sanford’s opinion that Ms. Borgsmiller is “unable to work” during her

flares of severe pain, Admin. R. at 158, is an opinion on the ultimate issue of

disability, which is a matter reserved to the Commissioner, 

, the ALJ gave the opinion great weight and thus accepted that Ms. Borgsmiller

cannot work during her flares of severe pain.

       But the ALJ omitted this limitation in her RFC assessment. As a result, the

ALJ’s RFC assessment contradicts her own determination that Ms. Borgsmiller is

unable to work during flares of severe pain. Further, while the ALJ stated in support

of her RFC assessment that Ms. Borgsmiller’s treating physicians did not place any

functional restrictions on her activities, this is accurate only with respect to those

periods when Ms. Borgsmiller is not experiencing flares of severe pain. When she is

experiencing flares of severe pain, Dr. Sanford opined that Ms. Borgsmiller is unable

to work, which represents a significant functional restriction on her activities.

Accordingly, the ALJ’s RFC does not accurately reflect Ms. Borgsmiller’s pain

limitation, and, therefore, the ALJ erred at the first phase of the step four analysis.

See, e.g., 

(noting ALJ’s error in failing to include in

RFC all the treating physician’s limitations that ALJ had credited).


                                          - 17 -
      Relatedly, we also find problematic the ALJ’s conclusion that “the flares do

not occur often enough to prevent [Ms. Borgsmiller] from engaging in substantial

gainful activity,” Admin. R. at 16. Ms. Borgsmiller testified that her flares last for

weeks, even months, at a time. The ALJ relied on evidence that the flares were

occurring with decreasing frequency in 2007 and 2008 to conclude that they did not

interfere with Ms. Borgsmiller’s ability to engage in substantial gainful activity.

      But even if the frequency of the flares had declined in 2007 and 2008, that

does not mean they were not disabling before 2007, when they occurred with more

frequency. Whether Ms. Borgsmiller’s flares happened with such frequency, and

were of such duration, as to prevent her from engaging in substantial gainful activity

between the November 2001 onset date and December 31, 2006, when she was last

insured for benefits, is a separate inquiry altogether. And it is one the ALJ did not

specifically address. We recognize that the ALJ need not discuss every piece of

evidence, Clifton v. Chater, 

, 1009-10 (10th Cir. 1996), and it is

possible that she considered the flares that occurred prior to 2007, in concluding that

they did not occur with sufficient frequency as to be disabling. But to the extent the

ALJ considered only the frequency of the flares that occurred after December 31,

2006, in reaching that conclusion, such consideration was erroneous.

      Ms. Borgsmiller also contends, and we agree, that the ALJ’s analysis at phase

three of step four was flawed. Phase three requires the ALJ to determine whether the

claimant has the ability to meet the job demands of the PRW found in phase two


                                         - 18 -
despite the mental and/or physical limitations found in phase one. 

. As noted previously, the ALJ failed to include Ms. Borgsmiller’s pain

limitation in the RFC assessment at phase one. She concluded, however, that

Ms. Borgsmiller could perform her PRW as a service office manager. [Admin. R.

at 16]. The ALJ’s finding regarding Ms. Borgsmiller’s ability to perform her PRW is

not supported by substantial evidence because it does not accurately reflect the record

or her own findings.

      In determining whether Ms. Borgsmiller is able to perform her PRW, the ALJ

sought information from the VE about Ms. Borgsmiller’s PRW. See 

(“[T]he ALJ may rely on information supplied by the VE at step four about

the demands of the claimant’s [PRW].”). The VE testified that Ms. Borgsmiller’s

PRW as a service office manager was most akin to a secretarial position, and

constituted a sedentary job, and that her PRW as an apartment manager constituted a

light job. Based on our review, the ALJ did not, however, adequately inquire about

the demands of Ms. Borgsmiller’s PRW in order to determine whether she could meet

those demands.

      Specifically, although the ALJ credited Dr. Sanford’s opinion that

Ms. Borsmiller is unable to work during flares of severe pain, the record does not

demonstrate that the ALJ sought information from the VE regarding whether the

demands of Ms. Borgsmiller’s PRW permitted absence from work for periods of time

of the length and frequency that the ALJ determined corresponded with


                                        - 19 -
Ms. Borgsmiller’s flares of severe pain. This error may have resulted, in part, from

the ALJ’s initial failure to make specific and adequate findings regarding the

frequency and duration of Ms. Borgsmiller’s flares of severe pain before 2007. 9 9F




Ms. Borgsmiller testified that there are “times [she] does not get out of bed for weeks

at a time,” Admin. R. at 32; see also 

she reported having been

bedbound “for up to 3 months at a time,” 

also 

there is

no evidence in the record demonstrating the vocational impact of these periods of

severe inactivity. Accordingly, the inadequacies of the ALJ’s inquiries to the VE

rendered her determination that Ms. Borgsmiller could perform her PRW

unsupportable.

      “In order to engage in [substantial] gainful activity, a person must be capable

of performing on a reasonably regular basis.” Byron v. Heckler, 

, 1235

(10th Cir. 1984). But without making necessary findings regarding the demands of

Ms. Borgsmiller’s PRW in conjunction with Ms. Borgsmiller’s pain limitation, the

ALJ could not, and did not, make an adequate finding regarding Ms. Borgsmiller’s

ability to perform her PRW on a regular and continuing basis. See 

(determination that claimant is able to engage in substantial gainful


9
       In order to make the necessary findings on remand, it may be necessary for the
ALJ to solicit more information from Ms. Borgsmiller concerning the frequency and
duration of her flares between her alleged onset date and December 31, 2006.




                                         - 20 -
activity requires a determination that claimant can hold whatever job she finds for a

significant period of time); 

. As such, the ALJ’s conclusion

that Ms. Borgsmiller may perform her PRW as a service office manager is not

adequately supported in the record.

       We conclude the ALJ erred in her step four analysis by failing to include

Ms. Borgsmiller’s pain limitation in the RFC assessment and in failing to make

inquiries of the VE necessary to determine the vocational impact of

Ms. Borgsmiller’s flares of severe pain. We, therefore, reverse and remand on this

issue also.

                                  III.   Conclusion

       The judgment of the district court is REVERSED, and this case is

REMANDED with instructions to remand to the agency for additional proceedings

consistent with the views expressed herein.


                                                  Entered for the Court


                                                  David M. Ebel
                                                  Circuit Judge




                                         - 21 -