I invite the interested reader to have a look at an article in the Journal of the American Medical Association, 300(13), pp. 1551-1565 (as of today, it is the current issue).
It presents some compelling evidence that long-term psychotherapy is significantly effective in treating a wide variety of psychiatric illnesses, including those illnesses that are more complicated, perhaps not neatly fitting into a single diagnostic category. It specifically shows a superior benefit for "psychodynamic" psychotherapy. Psychodynamic therapies are more intensive, philosophically and intellectually challenging styles of psychotherapy, which also emphasize the importance of examining closely the patient-therapist relationship during the therapy sessions. Most psychotherapy research has looked at short-term therapy (typically lasting a few months), and the evidence base formed from this research has caused a standard of practice to evolve which supports short-term but not long-term therapy. This is very problematic, because such a standard of practice is literally "short-sighted": Many patients require, and benefit greatly from, longer courses of psychotherapy. I do understand that it is difficult to do long-term research studies for any type of treatment (it is much, much easier to do an 8-week study than a 5-year study!), so this paper is very important.
Interestingly, the study did not show a specific benefit from the therapist having more years of experience. But this is not a surprise to me. In my experience with teaching medical students & residents, and also encountering many different supervisors during my training (which is now a decade ago), I found that differences in each individual's style, personality, attitude, and comfort with patients, were far more significant factors which influenced their helpfulness as clinicians or therapists, rather than years of experience. I can think of examples on both sides: the passage of years may magnify authoritarian or arrogant tendencies, and reduce open-mindedness, in some psychiatrists, particularly if they are ascending some sort of professional hierarchy, if they are trying to earn the most possible money, or if they are not enjoying their work. Clearly, this is not beneficial to patients or students. On the other hand, I have encountered many psychiatrists whose talent and wisdom have grown and deepened over time; in these cases I think years of experience have allowed a gentle, compassionate, understanding calm with patients, with a continuing open-mindedness to learn and continue growing, personally and professionally, as therapists and teachers.
The study did not show evidence that a "manual" helped the therapy be more effective. Again, this is absolutely not a surprise. Imagine trying to have a conversation with someone who was basing their responses to you on what a "manual" advised! While the use of manuals has been a well-meaning device for helping therapists-in-training to learn, and for measuring the effectiveness of therapy techniques in research, such devices in many cases can obtund the type of creative, spontaneous, friendly, wise, and sometimes playful interactions that I believe are the foundations for good therapy. These qualities I think are harder for therapists to learn or to teach, though I think that a sort of apprenticeship model is most likely to work best. It is not a surprise to me that the best such apprenticeship models are themselves long-term ones (over years), as opposed to having a therapist-in-training constantly migrating every few months, to some new or different therapy style or supervisor. Mind you, I do think it is important for any therapist to learn every major type of therapy, from a variety of different mentors, and in my opinion it is then necessary for the budding therapist to form an individual style that suits them best. In this way they will be truest to themselves, and therefore be best able to help their patients.