DECISION AND ORDER
LYNN ADELMAN, District Judge.Plaintiff Karen Hunt applied for disability insurance benefits and supplemental security income, claiming that she could no longer work due to mental health problems.1 (Tr. at 235, 238.) Denied initially (Tr. at 164-65, 188-91) and on reconsideration (Tr. at 166-67, 184-87, 192-95), plaintiff requested a hearing before an Administrative Law Judge (“ALJ”) (Tr. at 196), but the ALJ likewise found plaintiff not disabled and denied her applications (Tr. at 168-79). Plaintiff sought review from the Appeals Council (Tr. at 377), but the Council declined (Tr. at 1-3), making the ALJ’s decision the final word from the agency. See Arnett v. Astrue, 676 F.3d 586, 591 (7th Cir.2012). Plaintiff now seeks judicial review of that decision.
I. APPLICABLE LEGAL STANDARDS
A. Judicial Review
The court reviews an ALJ’s decision to ensure that he applied the correct legal standards and supported his decision with “substantial evidence.” Jelinek v. Astrue, 662 F.3d 805, 811 (7th Cir.2011). Substantial evidence is such relevant evidence as a reasonable mind could accept as adequate to support a conclusion. Schaaf v. Astrue, 602 F.3d 869, 874 (7th Cir.2010). While the court may not under this standard re-weigh the evidence or substitute its own judgment for that of the ALJ, it must nonetheless conduct a critical review of the record, ensuring that the ALJ adequately discussed the issues and built an accurate and logical bridge from the evidence to his conclusion. McKinzey v. Astrue, 641 F.3d 884, 889 (7th Cir.2011). Further, the court’s review is confined to the rationales offered by the ALJ, Shauger v. Astrue, 675 F.3d 690, 695-96 (7th Cir.2012); it may not affirm based on post-hoc justifications provided by the Commissioner’s lawyers, see, e.g., Spiva v. Astrue, 628 F.3d 346, 348 (7th Cir.2010). Finally, because the court need not defer to conclusions of law, if the ALJ commits legal error the court may reverse without regard to the volume of evidence in support of the factual findings. White ex rel. Smith v. Apfel, 167 F.3d 369, 373 (7th Cir.1999). Failure to comply with the Commissioner’s regulations and rulings for evaluating disability claims constitutes legal error. Prince v. Sullivan, 933 F.2d 598, 602 (7th Cir.1991).
B. Disability Standard
Disability is determined under a five-step sequential analysis. Weatherbee v. Astrue, 649 F.3d 565, 568-69 (7th Cir.2011) (citing Craft v. Astrue, 539 F.3d 668, 673 (7th Cir.2008)):
The first step considers whether the applicant is engaging in substantial gainful activity. The second step evaluates whether an alleged physical or mental impairment is severe, medically determinable, and meets a durational requirement. The third step compares the impairment to a list of impairments that are considered conclusively disabling. If the impairment meets or equals one of the listed impairments, then the applicant is considered disabled; if the im*1134pairment does not meet or equal a listed impairment, then the evaluation continues. The fourth step assesses an applicant’s residual functional capacity (RFC) and ability to engage in past relevant work. If an applicant can engage in past relevant work, [she] is not disabled. The fifth step assesses the applicant’s RFC, as well as [her] age, education, and work experience to determine whether the applicant can engage in other work. If the applicant can engage in other work, [she] is not disabled.
Craft, 539 F.3d at 674.
The claimant bears the burden of proof in each of the first four steps. Weatherbee, 649 F.3d at 569 (citing Briscoe ex rel. Taylor v. Barnhart, 425 F.3d 345, 352 (7th Cir.2005)). However, if she reaches the fifth step the burden shifts to the agency to present evidence establishing that the claimant possesses the RFC to perform other work that exists in a significant quantity in the national economy. Id. (citing Liskowitz v. Astrue, 559 F.3d 736, 740 (7th Cir.2009)). The agency may satisfy that burden by summoning a vocational expert (“VE”) to provide an assessment of the types of occupations in which claimants can work and the availability of positions in such occupations. Id. (citing Liskowitz, 559 F.3d at 743).
II. FACTS AND BACKGROUND
A. Plaintiffs Application and Supporting Materials
Plaintiff alleged a disability onset date of October 16, 2007 (Tr. at 235, 238), the date she was terminated from her warehouse worker job with Harley Davidson. According to the separation letter, Harley terminated plaintiff due to attendance problems and employee concerns regarding her bizarre behavior (and an object concealed beneath her jacket, which she refused to reveal but turned out to be a Bible). Police escorted plaintiff from the premises, and she was shortly thereafter admitted to a psychiatric hospital. (Tr. at 275, 295.)
In her disability report, plaintiff indicated that she was unable to return to work because of panic attacks, trouble concentrating, depression, crying spells, and paranoia around other people or unfamiliar surroundings. (Tr. at 307.) She reported a twelfth grade education, with no further vocational training and work experience in factory jobs. (Tr. at 308, 312, 314, 322.)
In her function report, plaintiff indicated that she cared for four children, including a disabled child with tantrums. She indicated that she could perform daily tasks “but with a delayed reaction in my own timing.” (Tr. at 344.) Her mother helped her with cooking, cleaning, and caring for the children. She indicated that her disability impeded her ability to react quickly, memorize, and remember. (Tr. at 345.) She listed hobbies of listening to music, playing organ and piano, and going to church. She sometimes did these things daily, sometimes she stayed in bed for days. (Tr. at 348.) She described problems dealing with adolescent children and co-workers. She indicated that she could not express her thoughts clearly under duress or depression; sometimes she felt dazed and had a difficult time recalling events. (Tr. at 349.)
In a third party function report, plaintiffs mother indicated that plaintiff did not sleep much because of depression and stress from dealing with her children. (Tr. at 331.) Plaintiff performed daily tasks and chores more slowly due to headaches, fatigue, and depression. (Tr. at 332-33.) Sometimes she experienced panic attacks, and plaintiffs mother would accompany her shopping because she forgot things. (Tr. at 334.) Plaintiffs moth*1135er also noticed unusual behaviors, such as picking at her face. (Tr. at 337.)
B. Medical Evidence
1. Treatment Records
The medical evidence includes records from several psychiatric admissions, both before and after the alleged disability onset date. On April 25, 2007, plaintiffs family brought her to the Milwaukee County Mental Health Complex (“MCMHC”), Psychiatric Crisis Service Center, stating that she was acting odd and out of control. (Tr. at 77, 83.) She presented as quiet and internally preoccupied and reported feeling sad. (Tr. at 77.) She denied any current medications. (Tr. at 78.) She was oriented to person only, her speech was slow and disorganized, and she was seen dancing in the waiting room. Her mood was sad, her affect blunted, and her thought process disorganized. (Tr. at 82.) She also displayed some bizarre gestures and was heard singing about the blood of Jesus. (Tr. at 84.) She was diagnosed with psychosis, NOS (not otherwise specified), and rule out schizophrenia, with a GAF2 of 20 (Tr. at 82), provided Zyprexa3 and Ativan4 (Tr. at 85), and then transferred to Aurora Psychiatric Hospital for further treatment (Tr. at 15, 18, 79, 86, 504).
At Aurora, Dr. Russell Temme diagnosed major depressive disorder with psychotic features versus brief psychotic reaction. Her admitting GAF was approximately 25-30, 50 on discharge. (Tr. at 15, 32, 501.) On her initial admission, plaintiff was restricted to the unit and placed on fifteen minute checks for safety. Plaintiffs mother indicated that plaintiff had been under increased stress because of problems with her children, recent job loss, and eviction from her home. (Tr. at 15, 19-20; 501, 505-06; see also Tr. at 21-24, 507-10.) Her mother further reported that plaintiff had been admitted to the MCMHC about nine years ago, but recovered and did well until this admission. (Tr. at 16, 502.)
For the first few days of her stay at Aurora, plaintiff appeared quite isolative, guarded, and somewhat disorganized. By April 28, she was noted to be much calmer, in control, and more focused. She tried to minimize her symptoms but appeared very depressed with blunted affect. She was encouraged and agreed to stay an additional day, despite wanting to go home. At the time of her April 29 discharge, plaintiff denied auditory or visual halluci*1136nations although she did admit hearing music when initially admitted. She was educated about medications, including Risperdal,5 which did help with her psychosis. She. was also advised to seek help from family in looking after her children while she rested and decreased her stress level. Given her depressed appearance, she was also started on Celexa, an anti-depressant. As she did not present as a danger to herself, she was allowed to leave in stable condition but against medical advice on April 29. (Tr. at 16, 502; see also Tr. at 513-19, Aurora Psychiatric Hospital progress notes.) She did agree to follow up with the partial hospitalization program on April 30, upon which out-patient appointments would be scheduled for medication management. (Tr. at 17, 35, 503.)
On October 16, 2007, plaintiff was again brought to the MCMHC, after she started chanting at work with her Bible in her coat, which caused her co-workers to become concerned. (Tr. at 68, 74, 388.) She appeared extremely sad, with flat affect, but denied suicidal or homicidal ideation. (Tr. at 68, 388.) She reported taking no medications aside from iron pills. (Tr. at 69.) She was admitted for observation for several hours (Tr. at 70, 71, 391), then discharged to Columbia St. Mary’s Hospital with a GAF of 40 (Tr. at 73, 76, 393, 396).
At Columbia St. Mary’s, Dr. Bruce Stevens assessed psychosis, NOS, with a GAF of 10 on admission, 60 on discharge. (Tr. at 401, 405.) She presented initially in an extremely psychotic state with pronounced religious delusions and tangential and flight of ideas in terms of thought processes. (Tr. at 404.) Dr. Stevens indicated that plaintiff was not very cooperative or forthcoming during the initial part of her hospitalization stay and lacked complete insight into the nature of her psychiatric disturbance. (Tr. at 401, 403, 404.) She was able to be convinced to stay in the hospital for further observation and evaluation, and she accepted psychotropic medications — Abilify6 and Depakote7 — as a trial for her psychosis. (Tr. at 401.) She demonstrated improvement in overall mood and degree of psychosis during her stay, and at the time of discharge she demonstrated no sign of psychosis or disturbance in mood. (Tr. at 401.) Dr. Stevens indicated that her prognosis was critically dependant upon her ability to follow through with out-patient psychiatric care and medication compliance. (Tr. at 402.)
Plaintiff commenced psychiatric treatment at Renew Counseling Services following her release from the hospital. On October 29, 2007, she saw a counselor, Natalie Has, LMFT. (Tr. at 413, 418, 485, 490.) On mental status exam, her general appearance was appropriate, but her mood was anxious and depressed. (Tr. at 416, 488.) Her manner of relating seemed helpless and needy, and her judgment and insight were impaired. She indicated that she used to think her daydreams were real but no longer did; she thought she was having hallucinations but realized they are not reality, just daydreams. (Tr. at 416, 488.) Her affect was labile, and her thought flow organized. (Tr. at 416, 488.) She reported severe difficulty sleeping, *1137moderate emotional lability, and moderate anxiety. (Tr. at 417, 489.) She further reported severe lethargy, severe difficulty concentrating, and moderate loss of interest in pleasurable activity. (Tr. at 417, 489.) She also reported severe crying spells, mild to moderate psychomotor retardation or agitation, and mild appetite change. (Tr. at 417, 489.) Ms. Klas diagnosed psychosis, NOS, with a current GAF of 41, recommending weekly individual therapy and a psychiatric consultation. (Tr. at 418, 490.)
Plaintiff returned to Ms. Klas on November 26, 2007, indicating that she had stopped taking her medications due to side effects. Klas noted little progress in treatment since the first visit. (Tr. at 412, 484.)
On December 6, 2007, Dr. Basil Jackson at Renew indicated that it was hard for plaintiff to give a straight narrative because of her tendency to become paranoid. (Tr. at 420, 483.) Dr. Jackson reviewed the records from St. Mary’s, including the indications of hallucinations, extreme anxiety, and religious preoccupations rising to the level of delusional thinking. (Tr. at 420, 483.) Despite this, plaintiff denied that she had any psychiatric problems or ever had any. (Tr. at 420, 483.) She was heavily involved in the Church of God in Christ, a Pentecostal denomination. (Tr. at 420, 483.) On exam, she appeared guarded but cooperative. (Tr. at 419, 482.) She spoke somewhat in a monotone, with flat affect. (Tr. at 419, 482.) There was some evidence of underlying depression and anxiety. She did not communicate enough verbally for it to be clear just how disorganized her thinking was or what delusions she might be experiencing. (Tr. at 419, 482.) While guarded, she was oriented and able to concentrate. (Tr. at 419, 482.) In the past she had been treated with Ability, and Dr. Jackson felt it appropriate to start her on that again. He also prescribed Trazodone8 for sleep. (Tr. at 419, 482.) He assessed bipolar disorder v. schizoaffective disorder. (Tr. at 419, 482.)
On February 21, 2008, Dr. Jackson found plaintiff guarded and rather agitated because she had to wait in the waiting room and felt people would tell her one thing and then tell her another thing. She did not want to take medicine. She was also upset because she had been turned down for social security, which Dr. Jackson did not understand, stating: “It’s clear she’s psychotic, paranoid.” (Tr. at 481.) She did agree to take medicine that would help her sleep, so Dr. Jackson continued the Trazodone and Ability, both at bedtime. Dr. Jackson noted that on this medicine plaintiff was able to sleep, and her mother noted that “as long as she sleeps she does reasonably well.” (Tr. at 481.) Dr. Jackson wrote: “At this time there’s no sign of abuse of medicine, side effects of medicine or suicidal ideation. But she is psychotic and very paranoid.” (Tr. at 481.)
On June 19, 2008, plaintiff returned to therapist Klas, who again noted little progress in treatment. (Tr. at 480.) On July 31, 2008, plaintiff told Klas that she felt depressed and unable to work. Klas noted greater insight into taking responsibility for mental health and did not observe any psychosis. (Tr. at 479.) Plaintiff also saw Dr. Jackson on July 31, 2008, asking for something to help her concentrate and focus. Dr. Jackson indicated that a psycho-stimulant would be the opposite of the medication she was taking, Ability. They agreed that because of her underlying de*1138pression, Flouxetine9 might help, and she agreed to take it. (Tr. at 475.)
Plaintiff returned to the clinic to see Klas on August 28, 2008, indicating that her symptoms remained the same. They discussed coping skills. (Tr. at 478.) On September 4, 2008, plaintiff told Klas that her depression inhibited her daily routine. (Tr. at 477.)
On September 4, 2008, Dr. Jackson completed a psychiatric impairment questionnaire, indicating that he had seen plaintiff six times since October 2007. He listed diagnoses of psychosis, NOS, and depression, with a current GAF of 41 and lowest in the past year also 41. (Tr. at 456.) He checked clinical findings of poor memory, oddities of thought, appetite disturbance, sleep disturbance, mood disturbance, social withdrawal, emotional lability, blunt affect, illogical thinking, delusions or hallucinations, persistent irrational fears, paranoia, and difficulty thinking or concentrating. (Tr. at 457.) He listed as symptoms flat affect, paranoia, disorientation, and delusions. (Tr. at 458.) He assessed marked limitation in her ability to understand and remember one or two step instructions, maintain attention and concentration for extended periods, and make simple work-related decisions; moderate limitation in the ability to remember locations and work-like procedures, understand and remember detailed instructions, carry out simple one or two step instructions, work in coordination with or proximity to others, complete a normal workday without interruptions from symptoms, interact with the public, get along with co-workers, respond appropriately to changes in the work setting, and set realistic goals; mild limitation in the ability to perform within a schedule, maintain attendance, ask simple questions, maintain socially appropriate behavior, and travel to unfamiliar places; and no limitation in the ability to be aware of normal hazards and take appropriate precautions. (Tr. at 459-61.) He indicated that she experienced episodes of deterioration or decompensation in the work setting, providing the example of her breakdown the previous year where she became preoccupied by religious thoughts. (Tr. at 461.) He opined that she was capable of moderate stress only; at past high stress jobs, she could only handle some of the situations. (Tr. at 462.) She would also experience “good days” and “bad days,” resulting in her being absent from work more than three times per month. (Tr. at 462-63.)
On September 6, 2008, plaintiff saw Klas, discussing her extreme problem with memory loss and concentration. Plaintiff reported going to school part-time, taking three college courses. (Tr. at 476.) Plaintiff returned to see Klas on January 7, 2009, indicating that she was doing good in school and focused on becoming a nurse. Klas noted that plaintiffs symptoms had decreased, and she was currently stable. Klas assessed good progress in treatment. (Tr. at 474.) Despite this progress, however, the record contains evidence of further trips to the MCMHC in January 2010 (Tr. at 59-67), and again in March 2010 (Tr. at 53-58) followed by an admission to Aurora Psychiatric Hospital with a GAF of 25 (Tr. at 37-41, 49).10
On April 15, 2010, Klas completed a psychological impairment questionnaire, indicating that she last saw plaintiff for *1139treatment on January 7, 2009. (Tr. at 492.) She listed diagnoses of psychosis, NOS, and depression, with a current GAF of 65, lowest in the past year of 41. (Tr. at 492.) She identified clinical findings of poor memory, appetite disturbance, sleep disturbance, decreased energy, generalized persistent anxiety, and difficulty thinking or concentrating. (Tr. at 493.) She listed symptoms of depression, anxiety, paranoia, poor memory, sleep disturbance, poor concentration, and disorientation. She indicated that plaintiff required hospitalization because of her condition in March 2010 and in 2007. (Tr. at 494.) She opined that plaintiff had marked limitations in the areas of understanding and remembering detailed instructions, maintaining attention and concentration for extended periods, performing within a regular schedule, working in coordination with or proximity to others, accepting instruction and criticism from supervisors, getting along with co-workers, responding appropriately to changes in the work setting, traveling to unfamiliar places, and setting realistic goals; moderate limitations in remembering locations and work procedures, understanding and remembering one or two step instructions, carrying out simple one or two step instructions, carrying out detailed instructions, making simple work-related decisions, completing a normal workday without interruptions from psychological symptoms, asking simple questions or requesting assistance, maintaining socially appropriate behavior, and being aware of normal hazards; and mild limitation in the ability to interact with the general public. (Tr. at 495-97.) Klas indicated that plaintiff previously experienced an episode of decompensation at work, resulting in her termination. (Tr. at 497.) IClas further indicated that plaintiff was capable of moderate stress. (Tr. at 498.) She would experience good and bad days, resulting in absences of more than three times per month or about once per month, depending on her cycle of mental health. (Tr. at 498-99.)11
2. SSA/State Agency Consultants
On January 4, 2008, James Paquette, Ph.D., performed a psychological evaluation for the SSA. (Tr. at 421.) He found plaintiff oriented x 3, courteous, and cooperative. Her attention/concentration was adequate, hygiene and dress age and weather appropriate, and speech and language abilities unremarkable. (Tr. at 421.) Depakote and Ability had been prescribed since October 2007, but she did not appear to be taking them on a regular basis due to adverse effects. (Tr. at 421-22.) She presented as pleasant and friendly, but reported loneliness and despondence, exacerbated by the recent separation from a boyfriend. She contacted a number of girlfriends during the week and socialized briefly with neighbors. (Tr. at 423.) Coping strategy review suggested that she employed withdrawal/isolation, and she reported no longer attending church. Her thought process was logical and goal-directed, and reality contact review asymptomatic concerning hallucinations, bizarre and non-bizarre delusions, or ideas of reference. (Tr. at 423.) Her mood was dysphoric with depressive and hypomanie symptoms, and she endorsed chronic despondence, discouragement, and frequent crying spells, but adequate appetite. Sleep pattern was compromised by nightly insomnia with racing thoughts. Poor concentration and inattention were noted, but *1140self-injurious behavior and suicidal ideation were denied. (Tr. at 423.) Her immediate memory was noted to be fair, recent memory fair, and remote recall adequate. Fund of information was also adequate. (Tr. at 423.) Abstract reasoning was intact, as was practical judgment/reasoning. Dr. Paquette diagnosed rule out adverse effects of medication, NOS; mood disorder, NOS; anxiety disorder, NOS, and rule out generalized anxiety disorder, with a GAF of 58. (Tr. at 424.) Concerning work capacities, Dr. Paquette found plaintiff able to understand, remember, and carry out simple instructions, and work adequately with co-workers and work supervisors. (Tr. at 424.)
On January 14, 2008, Jack Spear, Ph. D., prepared a psychiatric review technique report, evaluating plaintiff under Listings 12.03 (schizoaffective, paranoid and other psychotic disorders), 12.04 (affective disorders), and 12.06 (anxiety related disorders). (Tr. at 426.) Under the “B criteria” of the Listings, he assessed mild restriction of activities of daily living; mild difficulties in maintaining social functioning; moderate difficulties in maintaining concentration, persistence, and pace; and no episodes of decompensation. (Tr. at 436.) In an accompanying mental RFC report, Dr. Spear found marked limitations in the ability to understand, remember, and carry out detailed instructions; moderate limitation in the ability to maintain attention and concentration for extended periods, and respond appropriately to changes in the work setting; and no limitation in all of the other categories listed on the form. (Tr. at 440-41.)
On May 12, 2008, Pat Chan, M.D., completed a physical RFC report, finding no limitations other than avoidance of even moderate exposure to hazards (e.g., machinery, heights). (Tr. at 447-54.)
C. Hearing Testimony
On April 20, 2010, plaintiff appeared with counsel for her hearing before ALJ Patrick Toal. (Tr. at 96, 98.)
1. Plaintiff
On questioning by the ALJ, plaintiff testified that she used to go to church every Sunday, singing and playing the organ, but had not attended in the past year. (Tr. at 100-02.) She indicated that she had issues at her job at Harley Davidson because she was singing about Jesus (Tr. at 104), although her termination was also related to attendance problems (Tr. at 105). She testified that sometimes she did not recall certain periods of time and everything she did, and got “sick.” (Tr. at 105.) She stated that her problems were not physical, they were all mental health related. (Tr. at 105.) Asked to explain why she could not go to work, plaintiff said she did not “know how to explain it.” (Tr. at 106.)
Plaintiff testified that she had not worked since October 16, 2007, aside from a brief stint with a political campaign in 2008. (Tr. at 107.) She indicated that she had a driver’s license and sometimes drove a car but not during the times when she had “an episode” (Tr. at 108) — when she would go blank and not remember (Tr. at 109). She could not recall how often she had these episodes. (Tr. at 110.) She indicated that her condition worsened— and she was hospitalized — when she stopped taking her medication. She testified that she did not like taking her medications because the “side effects made me drool and stuff.” (Tr. at 110.) She admitted that she stopped taking the medication on her own without talking to her doctor. She also indicated that she quit taking her pills because she was teased — -“people label you.” (Tr. at 111.) She testified that when she took her medication she came “back to normal.” (Tr. at 112.) Pressed *1141by the ALJ to explain why she stopped taking her medication, she referred to the side effects and said: “I just don’t like the medicines.” (Tr. at 113.) She also worried that something could be wrong with her pills based on what she saw on TV. (Tr. at 113.) She indicated that when she stayed with her mother her mother gave her the medications, but she ordinarily lived by herself and did not take them. (Tr. at 114.)
Plaintiff testified that she took some courses at MATC but dropped out because she found it hard to keep up. (Tr. at 115-16.) She indicated that she wanted to go back after she got her medical condition under control. (Tr. at 117-18.) The ALJ then returned to plaintiff not taking her medications, soliciting an admission that if she did so she would be able to work and go to school. (Tr. at 119.) He further questioned her about why she missed appointments sometimes. (Tr. at 123-24.)
Plaintiff testified that she had four children, ages eleven, fifteen, sixteen and twenty, two of whom lived with her. (Tr. at 126-27.) She indicated that she was able to care for her children and participated in their education. (Tr. at 127.) She stated that she was able to leave her home, but her mother usually took her everywhere she went. (Tr. at 128.) She testified that she would like to work, “but the problem that have at work is when I go to work, they fire me.” (Tr. at 128.) Plaintiff indicated that she spent her time writing music. (Tr. at 134-36.)
Asked by her lawyer what was going through her head when she stopped taking her medications, plaintiff responded: “That I’m fine.” (Tr. at 137.) She also testified to concerns about the long-term side effects. (Tr. at 138.) Asked if she ever got “sick” when regularly taking her medications, plaintiff seemed uncertain but believed she had episodes in both circumstances. (Tr. at 141.) She also referred to issues with her insurance in obtaining medications. (Tr. at 141.) She had not, to her knowledge, required emergency room or crisis treatment when on her meds. (Tr. at 143.)
2. Vocational Expert
The VE, Ronald Raketti, classified plaintiffs past work as an order filler in the warehouse at Harley Davidson as light, semi-skilled work. (Tr. at 148-49.) Plaintiff also identified past work as a documentation coordinator at Waukesha Electric (Tr. at 152-53) and various factory jobs, including driving forklifts, operating metal presses and punch presses, and injection molding (Tr. at 155). After experiencing heart problems, she moved from factory to office work, including a stint as a receptionist/typist. (Tr. at 155-56.) Raketti classified the documentation clerk job as sedentary, semi-skilled; forklift operator as medium, semi-skilled; general office clerk as sedentary, semi-skilled; and machine operator as medium, semi-skilled. (Tr. at 156-57.)
The ALJ then asked a hypothetical question, assuming a person with plaintiffs educational and vocational background, exertionally unrestricted, but unable to maintain the attention and concentration necessary to perform detailed or complex tasks. The VE testified that such a person would not be able to perform plaintiffs past work, all of which was done at the semi-skilled level. (Tr. at 158.) However, the person could perform other jobs, including hand packaging, auto detailer, and janitor/cleaner. (Tr. at 159.) Adding to the hypothetical unscheduled breaks of undetermined duration and four or more absences per month would preclude all competitive employment. (Tr. at 161.) Unskilled employers would not tolerate *1142more than one unscheduled absence per month, twelve per year. (Tr. at 161.)
D. ALJ’s Decision
On May 28, 2010, the ALJ issued an unfavorable decision. (Tr. at 168.) Following the five-step process, the ALJ found that plaintiff had not engaged in substantial gainful activity since October 16, 2007, the alleged disability onset date, and that she suffered from the severe impairments of schizo/psychotic disorder, affective disorder, anxiety disorder, and migraine headaches, none of which met or equaled a Listing. (Tr. at 173.) The ALJ specifically considered Listing 12.04, finding under the B criteria of that Listing mild restriction of activities of daily living; mild difficulty in social functioning; moderate difficulty in concentration, persistence, and pace; and one episode of decompensation of extended duration. (Tr. at 174.)
The ALJ then determined that plaintiff retained the RFC to perform a full range of work at all exertional levels but was unable to maintain the attention necessary to perform complex tasks. (Tr. at 175.) In making this determination, the ALJ considered plaintiffs testimony, finding her statements “not credible to the extent they are inconsistent with the above residual functional capacity assessment.” (Tr. at 176.) The ALJ also considered Dr. Jackson’s report, finding it worthy of “some weight,” but not controlling. (Tr. at 177.) The ALJ further noted that the RFC he adopted was “in accord with the conclusions reached by physicians employed by the state agency.” (Tr. at 177.) The ALJ said nothing about therapist Klas’s report.
Based on this RFC, the ALJ concluded at step four that plaintiff was unable to perform any past relevant work. (Tr. at 177.) However, relying on the VE’s testimony, the ALJ determined at step five that plaintiff could perform other jobs existing in significant numbers. (Tr. at 178.) The ALJ therefore found plaintiff not disabled and denied her application. (Tr. at 179.)
III. DISCUSSION
Plaintiff argues that the ALJ erred in (1) considering the medical evidence from her treating professionals, (2) evaluating the credibility of her testimony, and (3) posing incomplete questions to the VE. I agree on all three points and therefore reverse and remand for further proceedings.
A. Consideration of Medical Evidence
1. Applicable Legal Standards
Social security regulations provide that more weight is generally given to opinions from treating physicians, as they are in the best position to provide a detailed, longitudinal assessment of the claimant’s condition. See, e.g., Farrell v. Astrue, 692 F.3d 767, 772-73 (7th Cir.2012) (citing 20 C.F.R. § 404.1527(c)). A treating source “may bring a unique perspective to the medical evidence that cannot be obtained from the objective medical findings alone or from reports of individual examinations, such as consultative examinations or brief hospitalizations.” 20 C.F.R. § 404.1527(c)(2). Accordingly, an ALJ must offer “good reasons” for discounting the opinion of a treating source. Scott v. Astrue, 647 F.3d 734, 739 (7th Cir.2011). If the opinion is “well-supported by medically acceptable clinical and laboratory diagnostic techniques” and “not inconsistent with the other substantial evidence” in the record, the ALJ must afford it “controlling weight.” 20 C.F.R. § 404.1527(c)(2); SSR 96-2p; Jelinek, 662 F.3d at 811; Punzio v. Astrue, 630 F.3d *1143704, 710 (7th Cir.2011). Even if the ALJ finds sound reasons for not giving the opinion controlling weight, he may not simply reject it, SSR 96-2p; rather, he must then decide what weight the opinion does deserve, considering a “checklist” of factors including the length, nature, and extent of the treatment relationship; frequency of examination; the physician’s specialty; the types of tests performed; and the consistency and support for the physician’s opinion. Scott, 647 F.3d at 740; Campbell v. Astrue, 627 F.3d 299, 308 (7th Cir.2010); see also Bauer v. Astrue, 532 F.3d 606, 608 (7th Cir.2008) (explaining that when the treating physician’s opinion is not given controlling weight “the checklist comes into play”). “In many cases, a treating source’s medical opinion -will be entitled to the greatest weight and should be adopted, even if it does not meet the test for controlling weight.” SSR 96-2p.
Opinions from non-physician providers, such as therapists, may not receive controlling weight. See 20 C.F.R. § 404.1513(d); SSR 06-03p. Nevertheless, opinions from these “other sources” are “important and should be evaluated on key issues such as impairment severity and functional effects, along with the other relevant evidence in the file.” SSR 06-03p; see also Barrett v. Barnhart, 355 F.3d 1065, 1067 (7th Cir.2004); Lauer v. Apfel, 169 F.3d 489, 494 (7th Cir.1999); Koschnitzke v. Barnhart, 293 F.Supp.2d 943, 950 (E.D.Wis.2003).
2. Analysis
a. Dr. Jackson’s Report
The ALJ erred in his evaluation of Dr. Jackson’s report. After summarizing the report, the ALJ wrote:
While Dr. Basil’s [sic] opinion is given some weight, it is not controlling. Dr. Basil had been treating the claimant for less than a year and had seen her only six times when he rendered this opinion. In addition, his opinion is inconsistent with other substantial evidence in the record, including evidence that she was taking three college courses in September 2008 and that her condition improved and stabilized when she took her medication as prescribed.
(Tr. at 177, record citation omitted.)
The lead reason the ALJ provided — the length and extent of Dr. Jackson’s treatment relationship with plaintiff — corresponds to the checklist, but the regulations make clear that the ALJ must first decide whether the opinion deserves controlling weight.12 As noted, controlling weight is given if the opinion is well-supported by medically acceptable clinical and laboratory diagnostic techniques and not inconsistent with the other substantial evidence. The ALJ provided no reason to discount *1144Dr. Jackson’s clinical technique and resultant findings, and, as discussed below, the supposed inconsistencies cited by the ALJ fail to provide a sound basis for rejection of the report.
It is true that in September 2008 plaintiff reported attending college part-time (Tr. at 476); in January 2009, she reported doing well in school, and therapist Klas noted that plaintiff was stable with decreased symptoms (Tr. at 474). However, the ALJ cited no evidence suggesting that plaintiff had improved sufficiently to work full-time. Attending college part-time is not equivalent to engaging in substantial gainful activity, see, e.g., Cohen v. Sec’y of Dept. of Health and Human Servs., 964 F.2d 524, 530 (6th Cir.1992), and the Seventh Circuit has rejected such reliance on “hopeful remarks” in treatment notes as a basis for discounting a treating source report, Bauer, 532 F.3d at 609. Nor is it appropriate to assume that a person characterized as “stable” is able to work. See, e.g., Hemminger v. Astrue, 590 F.Supp.2d 1073, 1081 (W.D.Wis.2008) (stating that “a person can have a condition that is both ‘stable’ and disabling at the same time”); Lechner v. Barnhart, 321 F.Supp.2d 1015, 1030 (E.D.Wis.2004) (“One can be stable and yet disabled.”).
Further, the ALJ failed to mention that plaintiff dropped out of school shortly thereafter, testifying at the hearing that she could not “keep up” (Tr. at 115), and that she wanted to return after she got her condition under control (Tr. at 117-18). The record also contains evidence of an additional psychiatric hospitalization in March 2010.13 (Tr. at 37-41, 53-58, 494.) This points up the danger of basing a decision on the claimant’s mental condition at a particular point in time, rather than considering the entire record. See Punzio, 630 F.3d at 710 (“As we have explained before, a person who suffers from a mental illness will have better days and worse days, so a snapshot of any single moment says little about her overall condition.”); see also Farrell, 692 F.3d at 773 (stating that “when a patient ... is only unpredictably able to function in a normal work environment, the resulting intermittent attendance normally precludes the possibility of holding down a steady job”).
Pervading the ALJ’s opinion — and his conduct of the hearing (see, e.g., Tr. at 118-19) — is the notion that if plaintiff would only take her pills regularly all would be well. However, the Seventh Circuit has recognized that mental illness “may prevent the sufferer from taking her prescribed medicines or otherwise submitting to treatment,” Kangail v. Barnhart, 454 F.3d 627, 630 (7th Cir.2006), a possibility the ALJ overlooked. See also Martinez v. Astrue, 630 F.3d 693, 697 (7th Cir.2011) (stating that “people with serious psychiatric problems are often incapable of taking their prescribed medications consistently”); White v. Comm’r of Soc. Sec., 572 F.3d 272, 283 (6th Cir.2009) (“For some mental disorders, the very failure to seek treatment is simply another symptom of the disorder itself.”); Pate-Fires v. Astrue, 564 F.3d 935, 945-46 (8th Cir.2009) (rejecting ALJ’s reliance on claimant’s non-compliance with medication because such non-compliance can be, and often is, the result of the mental impairment itself, which may deprive the claimant of the rationality to decide whether to continue treatment or medication); Nguyen v. Chater, 100 F.3d 1462, 1465 (9th Cir.1996) (noting that it is a questionable practice to chastise one with a mental impairment for *1145the exercise of poor judgment in seeking rehabilitation).
ALJs are also required to consider reasons why a claimant might not be complying with prescribed treatment. See Craft, 539 F.3d at 679. In the present case, plaintiff testified that she did not like taking her medications in part because of the side effects (Tr. at 110), a common concern in these types of cases. See, e.g., Spiva, 628 F.3d at 351 (“The administrative law judge’s reference to Spiva’s failing to take his medications ignores one of the most serious problems in the treatment of mental illness — the difficulty of keeping patients on their medications. The drugs used to treat schizophrenia, for example, can make a patient feel drowsy and stunned.”). Plaintiff also worried about the stigma of being labeled mentally ill (Tr. at 111), which courts have also recognized can be a valid reason for delaying or avoiding treatment. See, e.g., Sorn v. Barnhart, 178 Fed.Appx. 680, 681-82 (9th Cir.2006) (“Given the lingering social stigma of admitting to mental illness, Sorn’s explanation to her health care providers that she tried to put off seeking help as long as possible, out of shame — which the ALJ completely ignores — is entirely plausible.”). She also mentioned issues with her insurance (Tr. at 141), which may constitute a valid excuse for non-compliance. See Myles v. Astrue, 582 F.3d 672, 677 (7th Cir.2009).
In any event, the ALJ did not meaningfully consider whether plaintiffs condition was such that, even if she faithfully took her pills, she would be unable to work on a regular and sustained basis. As the Seventh Circuit recently stated in addressing a similar flaw in an ALJ’s decision: “The ALJ apparently concluded that Jelinek’s symptoms would have remained under control but for an unwillingness to take her medications as directed. But we have often observed that bipolar disorder, one of Jelinek’s chief impairments, is by nature episodic and admits to regular fluctuations even under proper treatment.” Jelinek, 662 F.3d at 814.
The ALJ noted that the RFC he adopted was in accord with the conclusions reached by the state agency physicians, presumably a reference to Dr. Spears’s report. (Tr. at 177.) However, an “ALJ can reject an examining physician’s opinion only for reasons supported by substantial evidence in the record; a contradictory opinion of a non-examining physician does not, by itself, suffice.” Cudgel v. Barnhart, 345 F.3d 467, 470 (7th Cir.2003).14
Therefore, the matter must be remanded for reconsideration of Dr. Jackson’s report. If the ALJ finds that the report is not entitled to controlling weight, he must evaluate it under all of the factors in the checklist.
b. Therapist Klas’s Report
The ALJ never mentioned Klas’s report. Although this report cannot receive controlling weight, it was legal error for the ALJ to ignore it entirely. See SSR 06-03p; see also Whitney v. Schweiker, 695 F.2d 784, 788 (7th Cir.1982) (“[A]n ALJ must weigh all the evidence and may not ignore evidence that suggests an opposite conclusion.”).
The Commissioner argues that Klas’s report does not support plaintiffs claim, providing several reasons why it should be rejected. (Def.’s Br. at 9-10.) However, the ALJ provided none of these reasons, *1146and judicial review is limited to the rationale set forth in the ALJ’s opinion. E.g., Spiva, 628 F.3d at 353.
Although he does not explicitly say so, the Commissioner may in making these arguments be appealing to the harmless error rule. The doctrine of harmless error is applicable to judicial review of administrative decisions, but the court will apply it only if “it is predictable with great confidence that the agency will reinstate its decision on remand because the decision is overwhelmingly supported by the record though the agency’s original opinion failed to marshal that support.” Id. None of the Commissioner’s arguments support application of the harmless error doctrine here.
The Commissioner first notes that Klas had not seen plaintiff for over a year at the time she prepared the report. However, the record does reflect a significant prior treatment history, and Klas’s findings were generally consistent with Dr. Jackson’s. The Commissioner further notes that Ellas assigned a current GAF score of 65, indicative of mild symptoms, which, he contends, is contrary to the severe limitations set forth in the report. However, Klas also assessed a lowest GAF of 41 within the past year, a score suggestive of severe symptoms. GAF scores often vacillate, see, e.g., Jelinek, 662 F.3d at 807 (noting that the claimant’s GAF scores fluctuated from a high of 65 to a low of 20 during hospitalization), and the Seventh Circuit has warned ALJs not to base RFC on a mentally ill claimant’s best days, Farrell, 692 F.3d at 773. Finally, the Commissioner notes that plaintiffs counsel solicited this report just before the hearing and suggests that Klas may have been trying to do plaintiff a favor. But “the fact that relevant evidence has been solicited by the claimant or her representative is not a sufficient justification to belittle or ignore that evidence,” Punzio, 630 F.3d at 712, and suspicions that a provider bent over backwards to assist a patient are properly considered as part of the “searching inquiry” required by the regulations, id. at 713. Ultimately, it will be up to the ALJ on remand to decide whether these arguments against Klas’s report undermine its persuasiveness. None of the Commissioner’s contentions support a finding that the report is so flawed as to make remand pointless.
B. Credibility
1. Applicable Legal Standards
In assessing the credibility of a social security claimant’s testimony, the ALJ must apply a two-step test. He must first decide whether the claimant suffers from a medically determinable impairment that could reasonably be expected to produce the symptoms alleged. SSR 96-7p. If the claimant suffers from no such impairments), or if the impairment(s) could not reasonably be expected to produce the claimant’s symptoms, the symptoms cannot be found to affect her ability to work. SSR 96-7p. If the ALJ finds that the claimant’s impairments) could produce the symptoms alleged, he must at the second step determine the extent to which the symptoms limit the claimant’s ability to work. SSR 96-7p. In making this determination, the ALJ must consider the entire record, including the claimant’s daily activities; the frequency and intensity of the symptoms; precipitating and aggravating factors; the type, dosage, effectiveness, and side effects of any medication the claimant takes; treatment, other than medication, for relief of symptoms; and any other factors concerning the claimant’s functional limitations and restrictions due to her symptoms. SSR 96-7p. The ALJ must then provide “specific reasons” for the credibility 'determination, supported by *1147the evidence and articulated in the decision. SSR 96-7p.
The reviewing court will give “an ALJ’s credibility determination special, but not unlimited, deference.” Shauger, 675 F.3d at 696. The ALJ must consider the factors set forth in the regulations, and he must support the credibility findings with evidence in the record. Id. Failure to do so constitutes error necessitating remand. See, e.g., Villano v. Astrue, 556 F.3d 558, 562 (7th Cir.2009).
2. Analysis
In evaluating credibility in the present case, the ALJ first employed a piece of “meaningless boilerplate seen frequently in decisions from ALJs,” Shauger, 675 F.3d at 696, stating:
After careful consideration of the evidence, the undersigned finds that the claimant’s medically determinable impairments could reasonably be expected to cause the alleged symptoms; however, the claimant’s statements concerning the intensity, persistence and limiting effects of these symptoms are not credible to the extent they are inconsistent with the above residual functional capacity assessment.
(Tr. at 176.) This boilerplate — the Commissioner has characterized it as a “template,” a passage drafted by the Social Security Administration for insertion into any opinion to which it pertains, Bjornson v. Astrue, 671 F.3d 640, 644-45 (7th Cir.2012) — -violates the SSR 96-7p requirement that the ALJ specify which statements are credible and which are not, explaining why by reference to all of the pertinent factors set forth in the regulations. See Spiva, 628 F.3d at 348. It also backwardly “implies that the ability to work is determined first and is then used to determine the claimant’s credibility,” Bjornson, 671 F.3d at 645, rather than evaluating credibility as an initial matter in order to come to a decision on the ultimate question of work capacity, Brindisi ex rel. Brindisi v. Barnhart, 315 F.3d 783, 788 (7th Cir.2003); see also Hildebrandt v. Astrue, No. 11-C-0739, 2012 WL 1571502, at *10 (E.D.Wis. May 3, 2012); Hayes v. Astrue, No. 11-C-0212, 2011 WL 4336754, at *7 (E.D.Wis. Sept. 14, 2011); Hoffman v. Astrue, No. 10-C-1152, 2011 WL 3236176, at *9 (E.D.Wis. July 27, 2011).15
The Commissioner notes that use of this boilerplate need not be fatal so long as the ALJ does more to explain his reasoning. See Richison v. Astrue, 462 Fed.Appx. 622 (7th Cir.2012). The ALJ did not do enough here. In addressing plaintiffs claims of disabling mental limitations, the ALJ stated:
The medical records indicate that the claimant functions well when she takes her medication as prescribed. Her condition stabilized to the point that she was able to take college courses when she was participating in individual therapy and taking her medication in 2008. During her testimony, the claimant complained of significant symptoms including severe memory loss and inability- to concentrate. However, she also admitted that she has not taken, her medication as prescribed because she is concerned about side-effects. Yet, there is no evidence that she has attempted to work with her doctor to adjust or change the medication in an effort to reduce or eliminate the side effects.
(Tr. at 177, record citations omitted.)
As discussed above, persons with mental illness often have difficulty complying *1148with a prescribed medication regimen, something the ALJ must consider “before concluding that non-compliance with medication supports an adverse credibility inference.” Jelinek, 662 F.3d at 814. As also discussed above, the ALJ’s reliance on plaintiffs brief improvement in late 2008 ignores the fact that plaintiff later quit school, hoping to get her condition under better control before re-enrolling. Even if plaintiffs condition at that time could be deemed indicative of lasting improvement, the ALJ made no finding that plaintiffs ability to attend school part-time meant she could work full-time.
The ALJ acknowledged plaintiffs complaint of side effects but faulted her for not resolving the issue with her doctor. However, this assumes that plaintiff had the capacity to rationally assess the situation and address her medication problems with her doctor. As discussed in the cases cited above, when dealing with severe mental illness, such an assumption cannot be made. The ALJ must take this into account when reconsidering credibility on remand.
C. YE Testimony
1. Applicable Legal Standards
The ALJ may rely on VE testimony in determining the availability of jobs the claimant can perform. See Liskowitz, 559 F.3d at 743. However, such reliance is proper only if the ALJ, in his hypothetical questions, oriented the VE to the totality of the claimant’s limitations, including any deficiencies in concentration, persistence, and pace. E.g., O’Connor-Spinner v. Astrue, 627 F.3d 614, 619 (7th Cir.2010); Stewart v. Astrue, 561 F.3d 679, 684 (7th Cir.2009); Young v. Barnhart, 362 F.3d 995, 1003 (7th Cir.2004). The Seventh Circuit has acknowledged some exceptions to this general rule, O’Connor-Spinner, 627 F.3d at 619-20, but in most cases the ALJ must “refer expressly to limitations on concentration, persistence and pace in the hypothetical in order to focus the VE’s attention on these limitations and assure reviewing courts that the VE’s testimony constitutes substantial evidence of the jobs a claimant can do,” id at 620-21.
2. Analysis
Plaintiff first argues that the hypothetical upon which the ALJ relied failed to include limitations set forth in the Jackson and Klas reports, including that plaintiff would be absent more than three times per month, a level of absenteeism the VE said would preclude work. (Tr. at 161.) The ALJ must on remand reconsider these reports, make any necessary modifications to the RFC determination, and, if warranted, solicit additional testimony from a VE.
Plaintiff further argues that the ALJ’s hypothetical did not even include all of the mental limitations recognized by the ALJ in his decision. Specifically, the ALJ found moderate difficulties in concentration, persistence, and pace (Tr. at 174), yet the only limitation he included in the hypothetical was that the person would be “unable to maintain the attention and concentration necessary to perform detailed or complex tasks.” (Tr. at 158.) In Stewart, the Seventh Circuit found that limiting a claimant to “simple, routine tasks” did not adequately account for documented limitations of concentration, persistence, or pace. 561 F.3d at 684-85; accord Young, 362 F.3d at 1004; see also O’Connor-Spinner, 627 F.3d at 618-19 (limiting claimant to “routine, repetitive tasks with simple instructions” did not account for “moderate limitation on concentration, persistence and pace”); Craft, 539 F.3d at 677-78 (limiting hypothetical to simple, unskilled work did not account for the claimant’s difficulty with memory, concentration, or mood *1149swings). In the present case, the ALJ essentially did the same thing, though he phrased the hypothetical in negative terms (i.e., the person cannot perform complex tasks). However, the Commissioner does not contend that this alternate phrasing triggers any of the exceptions to the general rule discussed in O’Connor-Spinner, 627 F.3d at 619-20. Indeed, the Commissioner does not respond to this argument at all.
“The ability to stick with a given task over a sustained period is not the same as the ability to learn how to do tasks of a given complexity.” Id. at 620. The ALJ must on remand ensure that his hypothetical questions to the VE adequately address plaintiffs moderate limitation in concentration, persistence, and pace.
IV. CONCLUSION
Plaintiff requests that the decision be reversed and the matter remanded for an award of benefits. In the alternative, she asks that the matter be remanded for further proceedings consistent with the court’s decision.
A judicial award of benefits “is appropriate only if all factual issues have been resolved and the record supports a finding of disability.” Briscoe, 425 F.3d at 356. As discussed above, issues remain for resolution in this case, including the weight to be assigned the reports of Dr. Jackson and therapist Klas, proper evaluation of plaintiffs testimony under SSR 96-7p, and the phrasing of hypothetical questions to the VE. The case must be remanded so that the ALJ may address these issues in the first instance. See Spaulding v. Astrue, 702 F.Supp.2d 983, 996 (N.D.Ill.2010).
THEREFORE, IT IS ORDERED that the ALJ’s decision is REVERSED, and this matter is REMANDED for further proceedings consistent with this decision. The clerk is directed to enter judgment accordingly.
. She also alleged disability due to migraine headaches but no longer presses that issue.
."GAF” — the acronym for "Global Assessment of Functioning” — rates a person’s psychological, social, and occupational functioning. Set up on a 0-100 scale, scores of 91-100 are indicative of a person with no symptoms, while a score of 1-10 reflects a person who presents a persistent danger of hurting herself or others. Scores of 81-90 reflect "minimal” symptoms, 71-80 "transient” symptoms, 61-70 "mild” symptoms, 51-60 "moderate” symptoms, 41-50 "severe” symptoms, 31-40 some impairment in reality testing, 21-30 behavior considerably influenced by delusions or hallucinations, and 11-20 some danger of hurting self or others. Diagnostic and Statistical Manual of Mental Disorders 32-34 (4th ed. 2000).
. Zyprexa is used to treat the symptoms of schizophrenia (a mental illness that causes disturbed or unusual thinking, loss of interest in life, and strong or inappropriate emotions). It is also used to treat bipolar disorder (manic depressive disorder; a disease that causes episodes of depression, episodes of mania, and other abnormal moods). It is in a class of medications called atypical anti-psychotics and works by changing the activity of certain natural substances in the brain. http://www. ncbi.nlm.nih.gov/pubmedhealth/PMH 0000161/.
. Ativan (Lorazepam) is used to relieve anxiety. It works by slowing activity in the brain to allow for relaxation, http://www.ncbi.nlm. nih.gov/pubmedhealth/PMH0000560/.
. Risperdal, an anti-psychotic, is used to treat the symptoms of schizophrenia. http://www. ncbi.nlm.nih.gov/pubmedhealth/PMH 0000944/.
. Abilify is used to treat the symptoms of schizophrenia; it is also used with an antidepressant to treat depression when symptoms cannot be controlled by the antidepressant alone. http://www.ncbi.nlm.nih.gov/ pubmedhealth/PMH0000221/.
.Depakote is used alone or with other medications to treat certain types of seizures and mania. http://www.ncbi.nlm.nih.gov/ pubmedhealth/PMH0000677/.
. Trazodone is used to treat depression. It works by increasing the amount of serotonin, a natural substance in the brain that helps maintain mental balance, http://www.ncbi. nlm.nih.gov/pubmedhealth/PMH0000530/.
. Fluoxetine (Prozac) is used to treat depression, obsessive-compulsive disorder, and panic attacks. http://www.ncbi.nlm.nih.gov/ pubmedhealtb/PMH0000885/.
. It appears that plaintiff's counsel forwarded the records related to the 2010 admissions on May 3, 2010 (Tr. at 14, 52), before the ALJ issued his decision, but it is unclear whether the ALJ saw them.
. The record also contains evidence and a treating source report related to plaintiffs migraine headaches. (Tr. at 465-66, 470-72; see also Tr. at 385-87.) Plaintiff makes no argument related to her headaches, so I do not further discuss that condition or the ALJ’s handling of it.
. In any event, the ALJ failed to explain why six visits over the course of nearly a year failed to provide Dr. Jackson with a valid basis for assessing plaintiff's condition. The ALJ’s reliance on this factor is particularly improper given his apparent crediting of the state agency physicians, who saw her once or not at all. In his brief, the Commissioner argues that because Dr. Jackson had treated plaintiff for just eleven months at the time he prepared the report, it is not clear that plaintiff's condition as described by Dr. Jackson met the durational requirement. See 20 C.F.R. § 404.1509 ("Unless your impairment is expected to result in death, it must have lasted or must be expected to last for a continuous period of at least 12 months. We call this the duration requirement.”). However, the ALJ did not discount the report based on the durational requirement, and judicial review is confined to the reasons he provided. Steele v. Barnhart, 290 F.3d 936, 941 (7th Cir.2002). In any event, Dr. Jackson specifically stated that plaintiff's impairments were ongoing, creating an expectation that they would last at least twelve months. (Tr. at 462.)
. While the ALJ may not have seen the records from this hospitalization, it was referenced in therapist Klas’s report, which clearly was before him, but which he ignored.
. Earlier in his decision, the ALJ discussed Dr. Paquette's psychological evaluation, but he did not formally credit that report. (Tr. at 176.) Even if he had, given the other flaws discussed in the text the case would have to be remanded for reconsideration of Dr. Jackson’s report.
. As I noted in Hayes, "There is nothing wrong with plugging into a decision boilerplate statements of law; doing so saves time for busy ALJs (and district judges).” 2011 WL 4336754, at *7 n. 14. But when an ALJ uses a "template” as a substitute for a specific credibility finding, he violates SSR 96-7p.