Patient satisfaction could be one outcome. The clinical outcomes might be only one factor for satisfaction. But as a whole the waiting time, out fo pocket expenditure, improvement of quality of life from a patient's perspective, and how they were treated would contribute for patient-centered outcomes.
To me (and from what I've heard PCORI-affiliated speakers say), patient-centered outcomes research must involve the patients and their perspective. As health care providers we can make assumptions about what is important, but those living with whatever condition we're looking at may have different ideas. For example, we're looking at disease severity by whatever categorization scale we have for it. For a patient, they're more concerned with pain, or frequency of medical visits, or fatigue, or whatever. It may not mean much if they go from level 3 to level 2, or remain at level 3 without progressing to level 4, but what is meaningful for them is their function and quality of life. Include patients in your research question development process.
Patient centered outcomes research means studying in terms of patient's compliance, acceptance, convenience and satisfaction regarding a health care service or modality of therapy.
In Evidence Based Medicine - clinical or individual level medicine - the patients preferences are one of three pillars guiding decisions. Patient preferences for informed choice need information.
In health system research or HTA it's not the perspective of the individual but of the society/patient group/health care system.
In research this different perspectives need different research questions especially with regard to outcomes.
On the health system level this is also shaping the representation of data - to a patientient centered data model.
Patient-centered outcomes research allows physicians to identify factors and treatments which most contribute to improving quality of life for our patients.
Using "satisfaction" and "quality of life, " both very subjective [and arguably meaningless] constructs, as part of the PCOR definition begs the question of what matters to patients. It's an n=1 question.
Optimal PCOR doesn't simply throw in a patient advisory board and a validated scale into the cauldron, but rather surfaces what is not already known: the domains, measures and thresholds that should, could or do drive the choices that patients, caregivers, and clinicians make. If PCOR isn't actionable, it's just navel-gazing.