Is it an efficient query when both keys examined & docs examined is high?

MongoDB Atlas
,
1

I’m trying to understand more on the slow query data log. In some of the slow queries, I saw some queries has both high “keys examined” & “docs examples” & “used indexes = true”. Does it suggest an efficient index-based query?

Here is an example of the query we have:

{
  "type": "command",
  "ns": "profiles.profiles",
  "command": {
    "find": "profiles",
    "filter": {
      "wallets.network": {
        "$exists": true,
        "$eq": "ETHEREUM"
      },
      "wallets.address": {
        "$exists": true,
        "$eq": "0x0000000000000000000000000000000000000000"
      }
    },
    "limit": 1,
    "singleBatch": true,
    "readConcern": {
      "level": "snapshot"
    },
    "lsid": {
      "id": {
        "$binary": {
          "base64": "G3/CI549RUmynPncgQoy9Q==",
          "subType": "04"
        }
      }
    },
    "txnNumber": 47622,
    "startTransaction": true,
    "autocommit": false,
    "$clusterTime": {
      "clusterTime": {
        "$timestamp": {
          "t": 1698949046,
          "i": 3
        }
      },
      "signature": {
        "hash": {
          "$binary": {
            "base64": "BnzSwRPGSbjI5v8SFjck/2kQ4AA=",
            "subType": "00"
          }
        },
        "keyId": 7237555922858410000
      }
    },
    "$db": "profiles",
    "$readPreference": {
      "mode": "primary"
    }
  },
  "planSummary": "IXSCAN { wallets.network: 1, wallets.address: 1 }",
  "planningTimeMicros": 117,
  "keysExamined": 16402,
  "docsExamined": 14633,
  "fromPlanCache": true,
  "nBatches": 1,
  "cursorExhausted": true,
  "numYields": 0,
  "nreturned": 0,
  "queryHash": "0D3949B0",
  "planCacheKey": "2E2ECD8F",
  "queryFramework": "sbe",
  "reslen": 231,
  "locks": {},
  "readConcern": {
    "level": "snapshot",
    "provenance": "clientSupplied"
  },
  "storage": {},
  "cpuNanos": 52353796,
  "remote": "3.134.126.8:2830",
  "protocol": "op_msg",
  "durationMillis": 105,
  "v": "7.0.2"
}
2

The difference between keys examined and doc examined is very small, so i think the used index is ok. Otherwise if the index is not good, then doc examined should be much higher.