Patients With a Condition - Video Solutions

Problem Overview: The table contains patient records with a conditions column storing multiple condition codes separated by spaces. The task is to return patients whose condition list contains the code DIAB1 as a standalone code, not as part of another code like DIAB100.

Approach 1: String Matching with Split Function (O(n * m) time, O(m) space)

Treat the conditions column as a space-separated list of codes. Split the string into tokens and check whether one of them equals DIAB1. The algorithm iterates through each patient record, performs a split(" ") operation, then scans the resulting list for the exact code. The key insight is avoiding substring matches like DIAB100 by comparing full tokens only. This approach is straightforward and works well when processing rows in application logic using languages like Python or JavaScript. Time complexity is O(n * m) where n is the number of rows and m is the average number of condition codes per row, with O(m) temporary space for the token list. This technique relies on basic string processing.

Approach 2: Using Regular Expressions (O(n * m) time, O(1) space)

Use a regular expression that matches DIAB1 as a full word inside the string. The pattern checks either the start of the string or a preceding space, followed by DIAB1, and ensures the code ends before another character continues the token. In SQL or backend filtering logic, a regex like (^|\s)DIAB1(\s|$) guarantees correct boundaries. Each row is scanned once by the regex engine, so the complexity remains O(n * m), but it avoids allocating intermediate arrays. This approach is cleaner when regex support is available in languages like Java or C#. It directly expresses the boundary condition and is common in database-style filtering problems involving regular expressions and SQL-style pattern matching.

Recommended for interviews: The regex approach is typically preferred because it expresses the "standalone code" requirement precisely and keeps the query concise. However, explaining the split-based method first shows you understand the underlying structure of the data and the risk of partial matches. Strong candidates often mention both approaches: tokenization for clarity and regex for compact production queries.