Harris (1969) reports outcomes of 39 mold arthroplasties for post-traumatic hip arthritis (acetabular fractures and dislocations) at MGH between 1945 and 1965. The paper simultaneously introduces a new 100-point hip function rating system designed to be more sensitive and broadly applicable than existing Larson and Shepherd systems. This instrument became the Harris Hip Score — the most widely used physician-administered hip outcome measure globally.
Harris weighted the score around surgical indication — pain and function dominate because they are the reason patients accept operative risk, while range of motion and deformity are nearly incidental:
Pain — 44 pts (single heaviest domain)
Function — 47 pts (gait 33 + daily activities 14)
Range of motion — 5 pts
Absence of deformity — 4 pts
Four result bands with precise thresholds must be memorized to read the literature accurately — a reported mean of 78 is fair, not a near-miss for good:
≥90 = excellent
80–89 = good
70–79 = fair
<70 = poor
The HHS distributed the same 39 hips across all four result bands (18 excellent, 7 good, 9 fair, 5 poor), while the Larson system rated 33/39 as good or excellent — compressing real functional differences and concealing clinical failure from the surgeon reading the data.
The deformity domain operates as an all-or-nothing penalty: any single finding — fixed flexion >30°, fixed adduction >10°, fixed internal rotation >10°, or limb-length discrepancy >3.2 cm — forfeits all 4 points, making a thorough contracture exam essential before scoring.
Outcomes improved significantly as technique matured — the second-half series (1955–1966) achieved 16/19 good or excellent results versus 8/18 in the first half (p < 0.05), attributed to the Aufranc concentric cup, refined acetabular reconstruction, better instruments, and improved postoperative care.
The HHS was not designed for total hip arthroplasty — it was built for young men with severe post-traumatic arthritis undergoing Smith-Petersen mold arthroplasty. Its heavy weighting toward pain and function reflects what drove those patients to surgery.
You need the score's structure cold for boards and for interpreting the literature. When a paper reports a mean postoperative HHS of 92, that is excellent by Harris's own thresholds. When it reports 78, that is fair. A result most patients and surgeons would not consider a success.
Know the deformity domain rule: all 4 deformity points are lost if the patient has a fixed flexion contracture >30°, fixed adduction >10°, fixed internal rotation >10°, or limb-length discrepancy >3.2 cm. Any single finding wipes the entire domain.
The HHS has a well-documented ceiling effect in primary THA populations. Patients cluster at the top, making it insensitive to differences between good and excellent outcomes. This is why modern trials pair the HHS with patient-reported measures (PROMIS, HOOS-Jr). The score is also examiner-administered, not patient-reported, which introduces bias when the surgeon scoring the result has a stake in it.
Harris (1969) reports outcomes of 39 mold arthroplasties for post-traumatic hip arthritis (acetabular fractures and dislocations) at MGH between 1945 and 1965. The paper simultaneously introduces a new 100-point hip function rating system designed to be more sensitive and broadly applicable than existing Larson and Shepherd systems. This instrument became the Harris Hip Score — the most widely used physician-administered hip outcome measure globally.
Harris weighted the score around surgical indication — pain and function dominate because they are the reason patients accept operative risk, while range of motion and deformity are nearly incidental:
Pain — 44 pts (single heaviest domain)
Function — 47 pts (gait 33 + daily activities 14)
Range of motion — 5 pts
Absence of deformity — 4 pts
Four result bands with precise thresholds must be memorized to read the literature accurately — a reported mean of 78 is fair, not a near-miss for good:
≥90 = excellent
80–89 = good
70–79 = fair
<70 = poor
The HHS distributed the same 39 hips across all four result bands (18 excellent, 7 good, 9 fair, 5 poor), while the Larson system rated 33/39 as good or excellent — compressing real functional differences and concealing clinical failure from the surgeon reading the data.
The deformity domain operates as an all-or-nothing penalty: any single finding — fixed flexion >30°, fixed adduction >10°, fixed internal rotation >10°, or limb-length discrepancy >3.2 cm — forfeits all 4 points, making a thorough contracture exam essential before scoring.
Outcomes improved significantly as technique matured — the second-half series (1955–1966) achieved 16/19 good or excellent results versus 8/18 in the first half (p < 0.05), attributed to the Aufranc concentric cup, refined acetabular reconstruction, better instruments, and improved postoperative care.
The HHS was not designed for total hip arthroplasty — it was built for young men with severe post-traumatic arthritis undergoing Smith-Petersen mold arthroplasty. Its heavy weighting toward pain and function reflects what drove those patients to surgery.
You need the score's structure cold for boards and for interpreting the literature. When a paper reports a mean postoperative HHS of 92, that is excellent by Harris's own thresholds. When it reports 78, that is fair. A result most patients and surgeons would not consider a success.
Know the deformity domain rule: all 4 deformity points are lost if the patient has a fixed flexion contracture >30°, fixed adduction >10°, fixed internal rotation >10°, or limb-length discrepancy >3.2 cm. Any single finding wipes the entire domain.
The HHS has a well-documented ceiling effect in primary THA populations. Patients cluster at the top, making it insensitive to differences between good and excellent outcomes. This is why modern trials pair the HHS with patient-reported measures (PROMIS, HOOS-Jr). The score is also examiner-administered, not patient-reported, which introduces bias when the surgeon scoring the result has a stake in it.