I like using printed D&D adventure modules. As a DM one usually doesn't have anybody to discuss with while preparing adventures; so if you just invent an adventure out of thin air, you don't have checks and balances to make sure you're not creating something which is boring, or unbalanced, or unbelievable. Using printed material and modifying it is more like a cooperative effort between me and the designer(s) of that adventure module, and thus usually gives a better result overall. Plus I profit from whatever handouts, maps, and other material are provided.

But I think the important part in this is to not use the modules as written, but to make the extra effort to modify them. One good reason for that is that a string of separate adventure modules does not make for a good campaign unless you put in various story hooks and connections to create a greater whole. Even the official D&D adventures that come as a series often have too little connections between adventures. But an even better reason for modifying adventures is getting the balance right.

4th edition D&D is not only a roleplaying game, but also a tactical miniature wargame. And it actually is quite a good tactical game at that if properly balanced. But the D&D module as written does not know how your group of players looks like. It makes an assumption of 5 players with at least one player of each role (tank, healer, dps, crowd control), which doesn't always correspond to reality. Thus if you as DM don't modify the encounters to take into account the strength of your actual group, you can quickly run into trouble.

I've been listening to Dungeons & Drogans I, a podcast of a DM and a group of players playing Keep on the Shadowfell. In session V of Dungeons & Drogans the players are doing the same kobold lair encounter my group did a while ago. Only my group was 6 players plus an NPC, while the group in the podcast only had 4 players, of which 2 were healers. So my group had an interesting and tough battle, while the Drogans basically wiped, in spite of some lucky critical rolls. Their DM had to pull out a Deux Ex Machina story device and offer the group to surrender and be held for ransom. The kobold wyrmpriest who had brought down the cleric even ended up healing him to prevent the character from dying.

The encounter as written is tough. It is encounter level 6, which is rather hard for a group of five level 1 characters, even if as written the second half of the kobolds only arrives with 3 rounds of delay. With my group of six plus one NPC I actually had to make the fight a bit harder by shortening that delay to just 1 round, which also made it somewhat more believable. But for a group of four players I would have removed some of the kobolds, at least one wyrmpriest and one of the kobold "tanks". I don't think the players in the podcast did anything wrong, they just simply weren't a match for the encounter as written.

Encounters in D&D serve two purposes: They provide a tactical game inside the roleplaying game, and they drive the story forwards. If the encounters are too easy, the story proceeds as planned, but the tactical game isn't all that interesting. If the encounters are too hard, the tactical game becomes frustrating, and the DM has to improvise to keep the story going, or admit defeat and kill the whole group. That sort of collective defeat (in D&D either everybody wins or everybody loses, including the DM) is something I would reserve for the players having done something actually really stupid. It isn't something that should happen because the DM failed to modify the encounter difficulty for the size of his group.

My advice: If you use a printed D&D adventure module, consider modifying the encounters if the size or composition of your group doesn't correspond to the assumptions of the writers. You can have a fun game of D&D with any size and composition of group, but it takes some preparatory work to get there.