

Human Centered Design Process
User Research
Stakeholder Analysis
Expert interviews
Benchmarking
Personas
Iterative Design
Dovetail
Figma
An investigation of existing barriers to accessing psychotherapy for individuals with public health insurance in Germany using the Human-Centered Design approach. And the development of a digital tool that improves access to psychotherapy for both psychotherapists and patients.
Access to psychotherapy in Germany is a structural problem with real consequences. Patients with statutory health insurance face long waiting times, fragmented information, and high organizational barriers. Therapists are simultaneously overwhelmed with inquiries and administrative work, with no efficient way to manage demand. Existing services like the "Terminservicestelle" and the "psychotherapeutische Sprechstunde" mitigate but do not solve the gap, especially for finding long-term therapy placements.
This Bachelor's thesis at THI Ingolstadt (graded 1.0) asked two research questions: what concrete barriers exist in the current placement process, and how a digital solution could improve it while respecting the needs of all stakeholders.
I applied the Human-Centered Design approach end-to-end.
Research happened on three levels:
I mapped the entire ecosystem: patients, therapists, "Terminservicestelle", statutory health insurers, "Bundespsychotherapeutenkammer", patient advocacy groups, and GPs. This clarified whose interests, constraints, and decision power shape the placement process.
(semi-structured, guideline-based) with nine participants across three perspectives:
This three-tiered framework was chosen deliberately:
I conducted a benchmark analysis of three existing digital tools: Doctolib, BetterHelp, and Tamly. In doing so, I analyzed their feature sets, user flows, and limitations in order to make design decisions based on the current market landscape.
Interview data was coded in two main categories: pain points and improvement options for patients and for therapists. These were synthesized into two evidence-based personas.
From there, I derived a structured requirements catalog:
This requirements layer became the bridge between research and design. Every prototype decision could be traced back to a specific finding.
The first prototype was a two-sided platform addressing both user groups in parallel.
Patient side: dashboard, therapy guide for first-time users, preferences flow, availability matching, therapist search with filters, therapist profiles, planning tool, FAQs, contact log, and curated resource links. This should help to not only search for therapists but also to be guided through the whole process while managing the contact attempts.
Therapist side: settings for appointment allocation, intake configuration, questionnaire management, opening hours, a calendar with confirmation flow, an inquiry inbox, and a waitlist with patient profile access. This should hepl to reduce administrative load without removing therapist control.
The prototype was tested in moderated user sessions with both groups. Findings were structured along strengths and problems for each side.
Patients responded well to the guided structure and the transparency of the search process. However, they flagged friction in the preferences flow and the therapy guide entry point.
Therapists valued the calendar and intake configuration, but the waitlist functionality and some settings paths needed rework.
Based on the test findings, key flows were redesigned in a second iteration:
Patients:
Therapists:
The thesis demonstrates that meaningful improvement in psychotherapy access is possible through design, but only when it respects the regulatory and operational constraints of the German healthcare system.
Some patient-side wishes, like a direct waitlist request function, are currently not legally feasible, which itself is a design finding worth surfacing. However, a guidance through the process is already a big improvement as patient's struggle to find and organize all the information themselves.
For therapists, a big improvement would be the structured handling of requests as they get so many that are manual handling is nearly impossible. Also structured and automated questionnaires can already reduce the manual workload a lot.