ChatGPT, the shoulder arthritis blog, and me

Many of us are curious about the utility of ChatGPT. I decided to put it to the test. I was recently asked if there was a way to summarize the 3000 posts on the Shoulder Arthritis Blog. I wondered whether ChatGPT could help.  Below is our conversation word for word. The whole conversation took less than 60 seconds.

Me: I have over 3000 posts on http://shoulderarthritis.blogspot.com can you help me summarize them by key word

ChatGPT: Here are some high-level key themes (keywords/topics) that emerge when reviewing your Shoulder Arthritis blog (with over 3,000 posts!):

🦴 Common Keywords / Topics

1. Reverse Total Shoulder Arthroplasty (rTSA)

Proposed Workflow for Summarization

Given the blog’s volume, here’s a scalable approach:

1. Content Harvesting

• Export all blog post titles, dates, labels (tags), and preferably the full text (via Blogger API or web scraping) into a structured dataset.

• Tokenization and keyword extraction (e.g., TF-IDF, RAKE).
• Topic modeling (LDA or BERTopic) to cluster posts into these themes.

• Count posts per topic over time (trend analysis).
• List most-used keywords and their co-occurrence networks.

• Generate concise summaries for each theme (e.g., top-5 posts, key clinical recommendations).
• Visualize trends with charts (volume by year, topic distribution).

• Provide an executive summary, keyword cloud, topic-trend graphs, and a way to drill down into specific themes/posts.

How I Can Help

I can assist you with:

• Designing the full pipeline (scraping, NLP, visualization).
• Writing scripts (in Python, R, etc.) to implement it.
• Suggesting tools you might use (e.g., Gensim, spaCy, BERTopic).
• Demonstrating a sample on a subset of posts.
• Generating summarized output, like an Excel or HTML summary.

Would you like to:

• Start with extracting titles/metadata?
• Dive into one theme in detail first (e.g., rTSA outcomes)?
• Or get a working prototype pipeline?

Let me know your preferred next step and I’ll tailor the workflow accordingly!

Me: rTSA outcomes

ChatGPT: Here’s a focused summary on reverse total shoulder arthroplasty (rTSA) outcomes based on the latest evidence:

📈 Functional Outcomes & Range of Motion

• Significant improvements in pain, shoulder function, and motion:
• Indications matter:

⚠️ Complication & Revision Rates

• Primary rTSA (≥2 yr follow-up):
• Mid to long-term (≥5 yr):
• Longer-term registry data:
• Scapular notching (grades I–II: 15–59%; III–IV: up to ~47%)
• Functional stability maintained in most, though some studies report deterioration or increased late revisions (up to 23% in select cohorts) pubmed.ncbi.nlm.nih.gov

🔁 Revision rTSA

• After failed primary rTSA:
• Revision from failed anatomic TSA:

🏃 Return to Work & Sport

🧩 Key Takeaways

1. Primary rTSA delivers strong pain relief and functional improvement, with sustained motion gains.
2. Complication and revision rates are modest short-term (~10% and <3%), increasing modestly over time.
3. Revisions are challenging, with higher complication risk but meaningful functional benefits.
4. Most patients resume sport and work, but recovery timelines depend on job/sport demands and psychosocial factors.

🔍 What’s Worth Monitoring

• High rates of scapular notching—does it influence long-term glenoid stability or implant survival?
• Long-term (>10 years) outcomes in larger registries as current data is limited.
• Predictors of successful return to activity beyond ROM—considering psychological and occupational factors.

I hope you found this response interesting.

Keeping and eye on opportunity

Osprey 

Union Bay Natural Area

Seattle

June 2025